Prediction method and device

Abstract

The invention discloses a prediction method and device, relates to the field of data processing, and is used for solving the problem of inaccurate inventory risk prediction. The method comprises the steps that the sales volume of a commodity is acquired, if the sales volume of the commodity is smaller than or equal to a first threshold value, a first distribution function is used for calculating the value of the commodity under a preset index, and if the sales volume of the commodity is larger than the first threshold value, a second distribution function is used for calculating the value of the commodity under the preset index, the first distribution function is a distribution function of which the probability of the first type of events is consistent, the first type of events are events of which the occurrence frequency is less than or equal to a first threshold value, the second distribution function is a distribution function of which the probability of the second type of events is consistent, and the occurrence frequency of the second type of events is greater than the first threshold value; the preset index is an index related to the inventory of the commodity, and predicting the inventory risk of the commodity according to the value of the commodity under the preset index. Through the scheme, the inventory risk of the commodity can be accurately predicted.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a prediction method and device. Background Art

[0002] To ensure effective inventory management and minimize losses, inventory metrics are often calculated to predict inventory risk and mitigate the risks of stockouts and overstocks. However, currently, regardless of the product, the primary metric used to measure inventory health is the stockout rate. This method of predicting inventory risk is inaccurate and can significantly differ from actual conditions.

[0003] Therefore, how to accurately predict the inventory risk of goods is an urgent problem to be solved. Summary of the Invention

[0004] The embodiments of the present invention provide a prediction method and apparatus for accurately predicting the inventory risk of commodities.

[0005] In a first aspect, an embodiment of the present invention provides a prediction method, comprising: obtaining the sales volume of a commodity; if the sales volume of the commodity is less than or equal to a first threshold, using a first distribution function to calculate the value of the commodity under a preset indicator; if the sales volume of the commodity is greater than the first threshold, using a second distribution function to calculate the value of the commodity under the preset indicator; wherein, the first distribution function is a distribution function that the probability of a first type of event conforms to, and the first type of event is an event whose number of occurrences is less than or equal to the first threshold; the second distribution function is a distribution function that the probability of a second type of event conforms to, and the second type of event is an event whose number of occurrences is greater than the first threshold; the preset indicator is an indicator related to the inventory of the commodity; and based on the value of the commodity under the preset indicator, the inventory risk of the commodity is predicted.

[0006] Through the above scheme, different distribution functions are used to calculate the inventory indicators of commodities according to the different quantities of commodities, thereby improving the accuracy of predicting commodity inventory risks and comprehensively evaluating the inventory status of commodities through multiple inventory indicators.

[0007] Optionally, the preset indicators include one or more of the following: out-of-stock rate indicator, near-expiry rate indicator, loss rate indicator without intervention; the first distribution function is a Poisson distribution function; or the second distribution function is a Gaussian distribution function.

[0008] The above solution can be used to estimate product inventory sufficiency, the proportion of products nearing expiration or nearing their expiration date within total inventory, and the potential for product loss nearing their expiration date. Furthermore, the Poisson and Gaussian distribution functions can better reflect the sales characteristics of different product types, thereby improving the accuracy of the estimates.

[0009] Optionally, the preset indicators include an out-of-stock rate indicator, and the first distribution function is used to calculate the value of the product under the preset indicators, including: predicting the first sales volume between the current moment and the moment of calculating the out-of-stock rate; substituting the first sales volume into the first distribution function, and calculating the value of the first distribution function at each fourth sales volume that is less than the beginning inventory based on the substituted first distribution function; obtaining the probability of sufficient inventory based on the value of the first distribution function at each fourth sales volume that is less than the beginning inventory, and determining the out-of-stock rate based on the probability of sufficient inventory.

[0010] Through the above scheme, the appropriate distribution function is selected according to the characteristics of different commodities, and factors such as the time of calculating the out-of-stock rate are included in the calculation, taking into account the uncertainty of commodity demand. Compared with traditional methods, the out-of-stock rate calculated by this method is closer to the actual situation.

[0011] Optionally, the preset indicators include an out-of-stock rate indicator, and the second distribution function is used to calculate the value of the product under the preset indicators, including: predicting the first sales volume between the current moment and the moment of calculating the out-of-stock rate; predicting the fluctuation value between the quantity of goods sold and the demand for the goods between the current moment and the moment of calculating the out-of-stock rate; substituting the first sales volume and the fluctuation value into the second distribution function, and calculating the probability of sufficient inventory based on the substituted second distribution function, and determining the out-of-stock rate based on the probability of sufficient inventory.

[0012] Through the above scheme, we select appropriate distribution functions based on the characteristics of different products, quantify the volatility of sales data, and conduct a comprehensive analysis while considering sales fluctuations, so as to estimate the out-of-stock rate more accurately.

[0013] Optionally, predicting the first sales volume between the current time and the time when the out-of-stock rate is calculated involves calculating the sales percentage based on the sales volume before the out-of-stock rate is calculated and the total sales volume for the current day. The first sales volume between the current time and the time when the out-of-stock rate is calculated is calculated based on the daily sales volume before the date when the out-of-stock rate is calculated and the product of the sales volume and the sales percentage on the date when the out-of-stock rate is calculated. This method considers the sales volume before the target day is calculated, rather than the total sales volume on the target day, making the calculation result closer to the actual out-of-stock rate at the time of observation.

[0014] Optionally, the preset indicators include an expiration rate indicator, and the first distribution function is used to calculate the value of the product under the preset indicators, including: predicting the second sales volume between the current moment and the day before the latest saleable date of the product, substituting the second sales volume into the first distribution function, and calculating the number of products that are about to expire the next day remaining in the warehouse one day before the saleable date of the products in the warehouse based on the substituted first distribution function, and determining the expiration rate based on the number of remaining products that are about to expire the next day and the beginning inventory.

[0015] Optionally, the preset indicators include an expiring rate indicator, and the second distribution function is used to calculate the value of the product under the preset indicators, including: predicting the second sales volume between the current moment and the day before the latest saleable date of the product, predicting the fluctuation value between the quantity of goods sold and the demand for the goods between the current moment and the day before the latest saleable date of the product, substituting the second sales volume and the fluctuation value into the second distribution function, and calculating the number of goods that are about to expire the next day remaining in the warehouse one day before the saleable date of the goods in the warehouse based on the substituted second distribution function, and determining the expiring rate based on the number of remaining expiring goods and the beginning inventory.

[0016] Optionally, the expiring rate is determined based on the quantity of the remaining expiring goods and the beginning inventory, including: determining the ratio of the quantity of the remaining expiring goods to the beginning inventory as the expiring rate, or obtaining the cost of the goods and the price of the goods; calculating the total cost of the expiring goods based on the quantity of the expiring goods and the cost of the goods; calculating the total price of the beginning inventory based on the beginning inventory and the price of the goods; and calculating the expiring rate based on the total cost of the expiring goods and the total price of the beginning inventory.

[0017] Optionally, the preset indicators include a loss rate indicator without intervention, and the first distribution function is used to calculate the value of the product under the preset indicators, including: predicting the third sales volume between the current moment and the saleable period of the product; substituting the third sales volume into the first distribution function, and calculating the loss quantity of the product in the warehouse based on the substituted first distribution function; determining the loss rate without intervention based on the loss quantity of the product and the beginning inventory.

[0018] Optionally, the preset indicators include a loss rate indicator without intervention, and the second distribution function is used to calculate the value of the product under the preset indicators, including: predicting the third sales volume between the current moment and the saleable period of the product; predicting the fluctuation value between the quantity of goods sold and the demand for the product between the current moment and the saleable period of the product; substituting the third sales volume and the fluctuation value into the second distribution function, and calculating the loss quantity of the goods in the warehouse based on the substituted second distribution function; determining the loss rate without intervention based on the loss quantity of the goods and the beginning inventory.

[0019] Optionally, determining the loss rate without intervention based on the number of lost products and the beginning inventory includes: determining the loss rate without intervention based on the ratio of the number of lost products to the beginning inventory; or obtaining the cost and price of the product; calculating the total cost of lost products based on the number of lost products and the cost of the product; calculating the total price of the beginning inventory based on the beginning inventory and the price of the product; and calculating the loss rate without intervention based on the total cost of lost products and the total price of the beginning inventory. In this way, taking into account the cost and price of the product makes the calculation result more accurate.

[0020] The opening inventory can be expressed as a decimal or integer. By taking into account non-integer inventory, inventory-related indicators can be calculated more accurately, reflecting the actual inventory situation more precisely and avoiding errors caused by rounding.

[0021] In a second aspect, an embodiment of the present invention provides a prediction device, comprising: an acquisition module for acquiring the sales volume of a commodity, a calculation module for using a first distribution function to calculate the value of the commodity under a preset indicator if the sales volume of the commodity is less than or equal to a first threshold, and using a second distribution function to calculate the value of the commodity under the preset indicator if the sales volume of the commodity is greater than the first threshold; wherein, the first distribution function is a distribution function that the probability of a first type of event conforms to, and the first type of event is an event whose number of occurrences is less than or equal to the first threshold; the second distribution function is a distribution function that the probability of a second type of event conforms to, and the second type of event is an event whose number of occurrences is greater than the first threshold; the preset indicator is an indicator related to the inventory of the commodity; a prediction module: for predicting the inventory risk of the commodity based on the value of the commodity under the preset indicator.

[0022] In one possible implementation, the calculation module is specifically used for: the preset indicators include one or more of the following: out-of-stock rate indicator, near-expiry rate indicator, loss rate indicator without intervention, the first distribution function is a Poisson distribution function; or, the second distribution function is a Gaussian distribution function.

[0023] In one possible implementation, the calculation module is specifically used to: preset indicators include an out-of-stock rate indicator, and use a first distribution function to calculate the value of the product under the preset indicators, including: predicting the first sales volume between the current moment and the moment of calculating the out-of-stock rate; substituting the first sales volume into the first distribution function, and calculating the value of the first distribution function at each fourth sales volume that is less than the beginning inventory based on the substituted first distribution function; obtaining the probability of sufficient inventory based on the value of the first distribution function at each fourth sales volume that is less than the beginning inventory, and determining the out-of-stock rate based on the probability of sufficient inventory.

[0024] In one possible implementation, the calculation module is specifically used to: preset indicators include an out-of-stock rate indicator, and use a second distribution function to calculate the value of the product under the preset indicators, including: predicting the first sales volume between the current moment and the moment of calculating the out-of-stock rate; predicting the fluctuation value between the number of products sold and the demand for the product between the current moment and the moment of calculating the out-of-stock rate; substituting the first sales volume and the fluctuation value into the second distribution function, and calculating the probability of sufficient inventory based on the substituted second distribution function, and determining the out-of-stock rate based on the probability of sufficient inventory.

[0025] In one possible implementation, the calculation module is specifically used to: predict the first sales volume between the current moment and the moment when the out-of-stock rate is calculated, including: calculating the sales volume before the moment when the out-of-stock rate is calculated and the sales volume for the whole day in a certain historical time period, and calculating the sales volume share; calculating the first sales volume between the current moment and the moment when the out-of-stock rate is calculated based on the daily sales volume before the date when the out-of-stock rate is calculated, and the product of the sales volume and the sales volume share on the date when the out-of-stock rate is calculated.

[0026] In one possible implementation, the calculation module is specifically used to: the preset indicators include an expiring rate indicator, and the first distribution function is used to calculate the value of the product under the preset indicators, including: predicting the second sales volume between the current moment and the day before the latest saleable date of the product, substituting the second sales volume into the first distribution function, and calculating the number of expiring products remaining in the warehouse one day before the saleable date of the products in the warehouse based on the substituted first distribution function, and determining the expiring rate based on the remaining number of expiring products and the beginning inventory.

[0027] In one possible implementation, the calculation module is specifically used to: preset indicators include an expiring rate indicator, and use a second distribution function to calculate the value of the product under the preset indicators, including: predicting the second sales volume between the current moment and the day before the latest saleable date of the product, predicting the fluctuation value between the quantity of goods sold between the current moment and the day before the latest saleable date of the product and the demand for the product, substituting the second sales volume and the fluctuation value into the second distribution function, and calculating the number of expiring goods remaining in the warehouse one day before the saleable date of the goods in the warehouse based on the substituted second distribution function, and determining the expiring rate based on the number of remaining expiring goods and the beginning inventory.

[0028] In one possible implementation, the calculation module is specifically used to: determine the near-expiry rate based on the number of remaining near-expiry goods and the beginning inventory, including: determining the ratio of the number of remaining near-expiry goods to the beginning inventory as the near-expiry rate, or obtaining the cost of the goods and the price of the goods; calculating the total cost of the near-expiry goods based on the number of near-expiry goods and the cost of the goods; calculating the total price of the beginning inventory based on the beginning inventory and the price of the goods; and calculating the near-expiry rate based on the total cost of the near-expiry goods and the total price of the beginning inventory.

[0029] In one possible implementation, the calculation module is specifically used to: preset indicators include a loss rate indicator without intervention, and use a first distribution function to calculate the value of the product under the preset indicators, including: predicting the third sales volume between the current moment and the product's saleable period; substituting the third sales volume into the first distribution function, and calculating the loss quantity of the product in the warehouse based on the substituted first distribution function; determining the loss rate without intervention based on the loss quantity of the product and the beginning inventory.

[0030] In one possible implementation, the calculation module is specifically used to: preset indicators include a loss rate indicator without intervention, and use a second distribution function to calculate the value of the product under the preset indicators, including: predicting the third sales volume between the current moment and the saleable period of the product; predicting the fluctuation value between the quantity of goods sold and the demand for the product between the current moment and the saleable period of the product; substituting the third sales volume and the fluctuation value into the second distribution function, and calculating the loss quantity of the goods in the warehouse based on the substituted second distribution function; determining the loss rate without intervention based on the loss quantity of the goods and the beginning inventory.

[0031] In one possible implementation, the calculation module is specifically configured to determine a no-intervention loss rate based on the number of lost items and the beginning inventory, including: determining the no-intervention loss rate based on the ratio of the number of lost items to the beginning inventory; or obtaining the cost and price of the item; calculating the total cost of lost items based on the number of lost items and the cost of the item; calculating the total price of the beginning inventory based on the beginning inventory and the price of the item; and calculating the no-intervention loss rate based on the total cost of lost items and the total price of the beginning inventory. In this way, taking into account the cost and price of the item makes the calculation result more accurate.

[0032] Optional, beginning inventory as a decimal or integer.

[0033] In a third aspect, the present application also provides a prediction device, which includes a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements the method described in the various possible designs of the first aspect.

[0034] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program or instruction is stored. When the computer program or instruction is executed by a processor, the method described in the various possible designs of the first aspect is implemented.

[0035] In a fifth aspect, an embodiment of the present application further provides a computer program product, which, when executed on a computer, enables the computer to execute the method described in the first aspect above.

[0036] These implementations or other implementations of the present application will be more concise and understandable in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 A schematic diagram of a flow chart of a prediction method provided by an embodiment of the present invention;

[0039] Figure 2 A schematic diagram of the relationship between the out-of-stock rate, the near-expiry rate, and the near-expiry loss rate without intervention provided in an embodiment of the present invention;

[0040] Figure 3 A schematic diagram of a flow chart for calculating out-of-stock rate when the quantity distribution of a commodity follows a Poisson distribution, provided by an embodiment of the present invention;

[0041] Figure 4 A schematic diagram of a flow chart for calculating out-of-stock rate when the quantity distribution of a product follows a Gaussian distribution, provided by an embodiment of the present invention;

[0042] Figure 5 A schematic diagram of a flow chart for calculating the near-expiry rate when the quantity distribution of a commodity follows a Poisson distribution, provided in an embodiment of the present invention;

[0043] Figure 6 A schematic diagram of a flow chart for calculating the near-expiry rate when the quantity distribution of a commodity follows a Gaussian distribution, provided in an embodiment of the present invention;

[0044] Figure 7 A schematic diagram of a flow chart for calculating loss rate without intervention when the quantity distribution of goods follows a Poisson distribution, provided in an embodiment of the present invention;

[0045] Figure 8 A schematic diagram of a flow chart for calculating loss rate without intervention when the quantity distribution of goods follows a Gaussian distribution, provided by an embodiment of the present invention;

[0046] Figure 9 A schematic structural diagram of a measuring device provided by an embodiment of the present invention;

[0047] Figure 10 A schematic structural diagram of another measuring device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solutions and beneficial effects of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0049] As mentioned in the background technology, currently, inventory status is calculated only by out-of-stock rate. This calculation method uses too simple an indicator, and the calculated results differ greatly from the actual inventory indicators, making it impossible to accurately estimate future inventory status. For example, the current method for calculating out-of-stock rate is usually:

[0050] 1) If the current inventory is less than the historical average sales volume, the out-of-stock rate Q is calculated as follows:

[0051]

[0052] 2) If the current inventory is greater than or equal to the historical average sales volume, the out-of-stock rate is 0, that is, Q = 0.

[0053] Because this calculation method fails to account for uncertainties in product demand, such as seasonal variations, market trends, and shifts in consumer behavior, out-of-stock rates calculated solely using this method are often underestimated and significantly different from actual conditions. Consequently, accurately estimating inventory indicators to predict product inventory risk is a pressing issue.

[0054] In view of this, the present invention provides a forecasting method, which uses different distribution functions to calculate inventory indicators of commodities with different sales volumes to solve the above problem.

[0055] Figure 1 A flow chart of a prediction method provided in an embodiment of the present invention includes the following steps:

[0056] Step 110: Obtain the sales volume of the product.

[0057] Step 120 : Determine whether the sales volume of the product is greater than a first threshold; if not, execute step 130 ; otherwise, execute step 140 .

[0058] Step 130: Calculate the value of the product under the preset indicator using the first distribution function.

[0059] The first distribution function is a distribution function that corresponds to the probability of a first type of event, and the first type of event is an event whose number of occurrences is less than or equal to the first threshold.

[0060] Step 140: Calculate the value of the product under the preset indicator using the second distribution function.

[0061] The second distribution function is a distribution function that corresponds to the probability of a second type of event, and the second type of event is an event that occurs more than the first threshold value.

[0062] Step 150: Predict the inventory risk of the product based on the value of the product under the preset indicators.

[0063] In a possible implementation, the first distribution function may be a Poisson distribution P(λ).

[0064] The Poisson distribution is a discrete probability distribution that describes the probability of an event occurring within a fixed time interval or spatial region. It is suitable for modeling rare events, particularly when the probability of an event is low and the events are independent of each other. The Poisson distribution has the following characteristics: 1. The events are independent, meaning that each event is independent of the occurrence of other events. 2. The events are rare, meaning that the probability of an event occurring is low but they are observed over a large time interval or spatial region. 3. The events have a constant average rate, meaning that the average rate of occurrence of an event is constant within a given time interval or spatial region.

[0065] The formula for the Poisson distribution P(λ) is as follows:

[0066]

[0067] Where k = 0, 1, 2, ...; λ is the average sales volume.

[0068] In a possible implementation, the second distribution function may be a Gaussian distribution N(μ,σ 2 ).

[0069] Gaussian distribution, also known as normal distribution, is a continuous probability distribution that describes the distribution of the sum of a large number of independent random variables. 2 ) is as follows:

[0070]

[0071] Based on the above, different distribution functions are used to calculate inventory indicators based on the number of products, improving the accuracy of inventory risk forecasts. This allows for a comprehensive assessment of inventory status across multiple inventory indicators. Furthermore, the Poisson and Gaussian distribution functions can better reflect the sales characteristics of different product types, thereby improving forecast accuracy.

[0072] In one possible implementation, inventory indicators may include one or more of the following: out-of-stock rate indicator, near-expiry rate indicator, or loss rate indicator without intervention. The following first introduces these three indicators and their respective uses:

[0073] Out-of-stock rates measure whether product inventory is adequate. A high out-of-stock rate can lead to lost sales opportunities, decreased customer satisfaction, and potential customer churn. A low out-of-stock rate can increase inventory costs and, for products with short shelf lives, increase shrinkage. Therefore, effectively managing out-of-stock rates requires balancing inventory levels with customer service quality, and is a key challenge in supply chain management.

[0074] The near-expiry rate is used to measure the proportion of goods that are about to expire or are close to their expiration date in the total inventory. This indicator is very important for managing inventory, reducing waste and optimizing procurement strategies, especially for products with limited shelf life, such as food and medicine. A high near-expiry rate will reduce customer satisfaction. For some special products (such as fresh products), a high near-expiry rate will also increase product losses. Therefore, measures need to be taken to speed up the sales or use of these products to avoid losses. A low near-expiry rate indicates that inventory management is more effective and product turnover is timely.

[0075] The no-intervention near-expiry loss rate measures the potential loss of goods as they approach their expiration date. The higher the no-intervention near-expiry loss rate, the greater the potential loss.

[0076] The above three inventory risk indicators are related to each other. For details, please refer to Figure 2 The relationship between the out-of-stock rate, the near-expiry rate, and the near-expiry loss rate without intervention is shown in the figure, where curve 1 is the changing trend of the near-expiry rate, curve 2 is the changing trend of the loss rate without intervention, and curve 3 is the changing trend of the out-of-stock rate. It can be seen in the figure that when the inventory of goods increases, the out-of-stock rate will decrease, while the near-expiry rate and the near-expiry loss rate without intervention will increase. Conversely, when the inventory of goods decreases, the near-expiry rate and the near-expiry loss rate without intervention will decrease, but the out-of-stock rate will increase. Therefore, it is necessary to balance the three indicators of out-of-stock rate, near-expiry rate, and loss rate without intervention to minimize inventory loss.

[0077] The following describes the calculation methods of out-of-stock rate, near-expiry rate, and loss rate without intervention through specific examples.

[0078] Before explaining, let's first introduce two important forms of warehousing in supply chain management: large warehouses and forward warehouses. Large warehouses usually refer to large warehouses located upstream of the supply chain, which are used to centrally store large quantities of goods. Large warehouses have a large area and can store large quantities of goods. Large warehouses usually serve as replenishment centers, providing goods for downstream forward warehouses, distribution centers and retail stores. Forward warehouses refer to small warehouses located downstream of the supply chain, close to consumer terminals. Due to their proximity to consumers, forward warehouses can achieve fast delivery and improve logistics timeliness. In summary, large warehouses are responsible for the centralized storage and replenishment of goods, while forward warehouses are close to consumers, improving delivery speed and flexibility.

[0079] Secondly, this example requires multiple data points, including but not limited to: ending inventory on the day before the target stocking date, estimated in-transit inventory to stores on the target stocking date, average sales volume at the forward warehouse, and standard deviation of demand. The following explains the data for different scenarios.

[0080] Scenario 1: Forecast sales and inventory in the near future (usually the next few days or weeks).

[0081] 1. Ending inventory on the day before the target stocking date A: Forecasted or estimated inventory level.

[0082] 2. Estimated in-transit inventory B on the target stocking date: This is the forecast or estimated quantity of merchandise that has been shipped from the supply source (e.g., the main warehouse) but has not yet arrived at the destination (e.g., the forward warehouse). This can be the quantity of merchandise that has a transfer plan (i.e., planned shipment from the main warehouse) but has not yet been received by the forward warehouse, or the quantity of merchandise that the forward warehouse is expected to receive without a transfer plan.

[0083] 3. Average sales volume PSD of the forward warehouse: sales volume predicted for the next few days (may be affected by activities, weather, holidays, etc.).

[0084] 4. Demand standard deviation σ1 (optional): The standard deviation between the predicted sales volume over a period of history and the actual sales volume when the product is not out of stock.

[0085] The purpose of short-term estimates is to more accurately assess recent inventory risk indicators, such as out-of-stock rate, near-expiry rate, and loss rate without intervention, etc., so as to help make more reasonable replenishment decisions.

[0086] Scenario 2: Use historical actual data to verify the accuracy of the forecasting model or method.

[0087] 1. Ending inventory A on the day before the target stocking date: actual historical inventory.

[0088] 2. Estimated in-transit inventory B that will arrive at the store on the target stocking date: The actual number of goods that have been shipped from the supply source (such as the warehouse) and arrived at the destination (such as the forward warehouse) on the target date.

[0089] 3. Average sales volume PSD for forward warehouses: This refers to the historical average actual sales volume for forward warehouses, or the historical average sales volume when stocks are always available. You can use historical daily actual sales volume / sales volume when stocks are always available. For example, for a given week, the average actual sales volume for each forward warehouse when stocks are always available is 2, 3, 4, 3, 2, 4, and 3, respectively.

[0090] 4. Demand standard deviation σ1 (optional): The standard deviation between the actual sales volume of a group of forward warehouses / historical sales volume when there is no shortage of goods and the average sales volume of the forward warehouses.

[0091] The purpose of historical validation is to evaluate the accuracy and reliability of the forecasting method by comparing the forecast results with actual historical data, which helps to improve the forecasting model and increase the accuracy of future forecasts.

[0092] Scenario 3: Forecast inventory and sales over a long period of time in the future.

[0093] For long-term forecasts, we assume that inventory will fluctuate around a target value. In other words, if the historical end-of-period inventory remains at a certain level, it will also fluctuate around this level in the future. This allows us to estimate future inventory levels.

[0094] 1. Ending inventory A on the day before the target stocking date: The theoretical inventory level, for example, the average safety stock is used as the ending inventory.

[0095] 2. Estimated in-transit inventory B on the target stocking date: This is the theoretical number of items that will arrive at the destination (e.g., the forward warehouse) on the target date. This can be based on the forward warehouse's average sales volume or a fixed transfer volume.

[0096] 3. Average sales volume PSD of the forward warehouse: This is the historical average sales volume of the forward warehouse (which can be categorized by holiday and non-holiday). For example, for a forward warehouse with weekly actual full-stock sales of 2, 3, 4, 3, 2, 4, and 3, the historical average sales volume is 3 units. For another example, the average sales volume from Monday to Friday is 2.8 units, and the average sales volume on weekends is 3.5 units.

[0097] 4. Demand standard deviation σ1 (optional): The standard deviation between the average sales volume over a historical period (which can be differentiated by holidays and non-holidays) and the actual sales volume when there is no shortage of stock.

[0098] The purpose of long-term forecasting is to evaluate and optimize long-term inventory strategies and provide guidance for future inventory management.

[0099] In addition, the following examples also require the use of opening inventory: opening inventory s = the previous day's ending inventory A + the estimated in-transit inventory B arriving at the store that day. That is, opening inventory s takes into account the previous day's remaining inventory and the new inventory that is about to arrive to accurately reflect the current actual inventory status.

[0100] The beginning inventory s does not necessarily need to be estimated using integers, so the decimal part of the beginning inventory s is taken as ω, that is, ω = s% 1. With a probability of ω, the beginning inventory is s (s rounded up), and with a probability of 1-ω, the beginning inventory is s (s is rounded down.) By taking non-integer inventory into account, inventory-related indicators can be calculated more accurately, reflecting the actual inventory situation more precisely and avoiding errors caused by rounding.

[0101] Similarly, the fractional part of the previous day's ending inventory A is taken as ω', that is, ω' = A% 1. With a probability of ω', the previous day's ending inventory is A (A is rounded up), and with a probability of 1-ω', the previous day's ending inventory is A (A is rounded down).

[0102] Based on the above content, the calculation methods of out-of-stock rate, near-expiry rate, and loss rate without intervention are explained in detail below.

[0103] Out-of-stock rate

[0104] There are multiple ways to calculate the out-of-stock rate, including but not limited to the following calculation methods 1 to 3.

[0105] Calculation method 1: Calculation of actual out-of-stock rate.

[0106] For a forward warehouse: the out-of-stock rate is calculated using the following formula:

[0107]

[0108] For multiple forward warehouses with similar characteristics within a region (or under the same supply warehouse), the out-of-stock rate is calculated using the following formula:

[0109]

[0110] Calculation method 2: Calculation of the out-of-stock rate when the product quantity distribution follows the Poisson distribution P(λ).

[0111] Here, the commodity quantity distribution obeys the Poisson distribution function, which means that the commodity quantity is less than or equal to the first threshold. In other words, the commodity can be considered as a medium-to-low-selling commodity.

[0112] Figure 3 This is a schematic diagram of the calculation process of the out-of-stock rate when the product quantity distribution follows the Poisson distribution. The specific steps are as follows:

[0113] Step 310: Calculate the sales volume before the out-of-stock rate and the sales volume for the entire day in a certain historical time period, and calculate the sales volume ratio.

[0114] Since the out-of-stock rate is a time-related indicator, for example, the out-of-stock rate calculated at 12 o'clock is different from the out-of-stock rate at 16 o'clock and 23 o'clock. Generally speaking (without replenishment), the later the out-of-stock rate is calculated, the higher the out-of-stock rate.

[0115] For example, if the out-of-stock rate is calculated at 16:00 on the afternoon of the replenishment day, the sales ratio sale_rate is calculated as follows:

[0116]

[0117] In this way, the sales volume before calculating the out-of-stock rate on the target day is taken into account instead of the sales volume for the entire day on the target day, so that the calculation result is closer to the out-of-stock rate situation at the actual observation time.

[0118] Step 320: Calculate the first sales volume between the current time and the time when the stock-out rate is calculated based on the daily sales volume before the date when the stock-out rate is calculated and the product of the sales volume and the sales volume percentage on the date when the stock-out rate is calculated.

[0119] The daily sales volume before the out-of-stock rate date can be calculated using accumulation, multiplication, power function, exponential function, etc. There are no restrictions here.

[0120] Optionally, the first sales volume is calculated as follows:

[0121]

[0122] Where i is the i-th day in the replenishment cycle, which is the day for evaluating the out-of-stock rate. When the replenishment cycle is 1 day, only the out-of-stock situation on the first day is considered, that is, i=1; target_date is the target stocking date; psd target_date+z-1 It is the daily sales volume before the date of out-of-stock rate calculation; psd target_date+i-1 Is to calculate the sales volume on the day of out-of-stock rate; sale_rate*psd target_date+i-1 It is the sales volume of the day before the out-of-stock rate is calculated (for example, before 16:00).

[0123] Step 330: Substitute the first sales volume into the Poisson distribution function, and calculate the value of the Poisson distribution function at each fourth sales volume that is less than the beginning inventory based on the substituted Poisson distribution function.

[0124] From the above introduction, we can know that the calculation method of Poisson distribution function is as follows:

[0125]

[0126] Wherein, λ is calculated in step 320. Substitute the λ calculated in step 320 into the formula of the Poisson distribution function to calculate the value of P(k).

[0127] Step 340: Obtain the probability of sufficient inventory based on the value of the Poisson distribution function at each fourth sales volume that is less than the initial inventory.

[0128] The values ​​of P(k) calculated in step 330 are accumulated to obtain the probability that the sales volume x in the warehouse is less than the initial inventory s, that is:

[0129]

[0130] Here, x is sales volume and s is beginning inventory.

[0131] Optionally, since the beginning inventory can be a decimal, from the above analysis, with a probability of ω, the beginning inventory is s (s rounded up), and with a probability of 1-ω, the beginning inventory is s (s is rounded down), so the probability that the sales volume x in the warehouse is less than the beginning inventory s can also be:

[0132]

[0133] Where cdf is the cumulative distribution function. This cumulative distribution function describes the probability that a random variable will be less than or equal to a specific value. In the above formula, cdf(s-ε) represents the probability that the random variable will be less than or equal to (s-ε), where ε is the minimum value.

[0134] Step 350: Determine the out-of-stock rate based on the probability of sufficient inventory.

[0135] P1=P(x≥s)

[0136] =1-P(x <s)

[0137] =1-cdf( s -1)-ω*P(x= s )

[0138] Where P1 is the out-of-stock rate.

[0139] Through the above scheme, the appropriate distribution function is selected according to the characteristics of different commodities, and factors such as the time of calculating the out-of-stock rate are included in the calculation, taking into account the uncertainty of commodity demand. Compared with traditional methods, the out-of-stock rate calculated by this method is closer to the actual situation.

[0140] Calculation method 3: The quantity distribution of goods follows Gaussian distribution N(μ,σ 2 ) in the case of out-of-stock rate.

[0141] Here, the commodity quantity distribution obeys the Gaussian distribution function, which means that the commodity quantity is greater than the first threshold. In other words, the commodity can be considered to be a high-selling commodity.

[0142] Figure 4 The quantity distribution of goods follows Gaussian distribution N(μ,σ 2 ), the flow chart of out-of-stock rate calculation is as follows:

[0143] Step 410: Calculate the sales share based on the total sales volume before the out-of-stock rate calculation time in a certain historical time period and the total sales volume for the entire day in the time period.

[0144] For example, if the product is replenished at 16:00 in the afternoon, the sales ratio sale_rate is calculated as follows:

[0145]

[0146] In this way, the sales volume before the out-of-stock rate is calculated is taken into account instead of the sales volume of the whole day on the out-of-stock rate calculation date, so that the calculation result is closer to the out-of-stock rate situation at the actual observation time.

[0147] Step 420: Calculate the first sales volume between the current time and the time when the stock-out rate is calculated based on the daily sales volume before the date when the stock-out rate is calculated and the product of the sales volume and the sales volume percentage on the date when the stock-out rate is calculated.

[0148] The daily sales volume before the out-of-stock rate calculation date can be calculated using accumulation, multiplication, power function, exponential function, etc., without any restrictions.

[0149] Optionally, the first sales volume is calculated as follows:

[0150]

[0151] Where i is the i-th day in the replenishment cycle. When the replenishment cycle is 1 day, only the out-of-stock situation on the first day is considered, that is, i=1; target_date is the target stocking date; psd target_date+z-1 It is the daily sales before the date of out-of-stock rate calculation; psd target_date+i-1 The out-of-stock rate is calculated by the sales volume for the whole day; sale_rate*psd target_date+i-1 It is the daily sales volume before the out-of-stock rate is calculated (for example, before 16:00).

[0152] Step 430: Predict the fluctuation between the quantity of goods sold and the demand for the goods between the current time and the time when the out-of-stock rate is calculated.

[0153] Current methods for calculating inventory indicators often ignore the volatility of product demand, such as seasonality, market trends, and changes in consumer behavior. This results in an inability to accurately estimate inventory risk. Therefore, this embodiment considers the fluctuation between the quantity of goods sold and the demand for the goods, resulting in more accurate calculation results.

[0154] Optionally, the fluctuation between the number of units sold and the quantity demanded can be calculated using the standard deviation. The standard deviation is a metric used in statistics to measure the degree of dispersion of a set of data. It represents the average distance between the data and its mean (i.e., how much the data fluctuates relative to the mean). The smaller the standard deviation, the more concentrated the data is around the mean; the larger the standard deviation, the more dispersed the data is.

[0155] Optionally, the standard deviation σ of the target time on day i in the replenishment cycle is calculated as follows:

[0156]

[0157] Where σ1 is the standard deviation of demand.

[0158] Step 440: Substitute the first sales volume and the fluctuation value into the Gaussian distribution function, and calculate the probability of sufficient inventory based on the substituted Gaussian distribution function;

[0159] From the above introduction, we can know that the calculation method of Gaussian distribution function is as follows:

[0160]

[0161] Where μ is calculated in step 420, and σ is calculated in step 430. Substituting λ calculated in step 320 and σ calculated in step 430 into the Gaussian distribution formula, we can obtain the probability that the sales volume x in the warehouse is less than the initial inventory s, that is:

[0162] p(x <s)=cdf(s-ε)

[0163] Where x is sales volume and s is beginning inventory.

[0164] The probability that the sales volume x in the warehouse is less than the beginning inventory s can be:

[0165] p(x <s)

[0166] =cdf(s-ε)

[0167] =cdf(s)

[0168] Step 450: Determine the out-of-stock rate based on the probability of sufficient inventory.

[0169] P1=P(x≥s)

[0170] =1-P(x <s)=1-cdf( s )

[0171] Among them, P1 is the out-of-stock rate.

[0172] Through the above scheme, we select appropriate distribution functions based on the characteristics of different products, quantify the volatility of sales data, and conduct a comprehensive analysis while considering sales fluctuations, so as to estimate the out-of-stock rate more accurately.

[0173] Premature Term Rate

[0174] There are many ways to calculate the near-term rate, including but not limited to the following calculation methods 1 and 2.

[0175] Calculation method 1: Calculation of the expiration rate when the quantity distribution of goods follows the Poisson distribution P(λ).

[0176] Here, the commodity quantity distribution obeys the Poisson distribution function, which means that the commodity quantity is less than or equal to the first threshold. In other words, the commodity can be considered as a medium-to-low-selling commodity.

[0177] Figure 5 This is a flowchart for calculating the expiration rate when the quantity distribution of goods follows the Poisson distribution P(λ). The specific steps are as follows:

[0178] Step 510: Predict the second sales volume between the current time and the day before the latest available sale date of the product;

[0179] The average remaining number of units between the current time and the day before the latest saleable date of the product can be predicted using accumulation, multiplication, power function, exponential function, etc., without any restrictions here.

[0180] Optionally, the first sales volume is calculated using the following formula:

[0181]

[0182] Among them, sp is the saleable period of the product, that is, the maximum time during which the sale can occur. For example, some aquatic products are only sold on the same day, so sp = 1. For example, some root vegetables have a saleable period of sp = 3, which means that the product can be sold for a total of 3 days after it arrives at the forward warehouse.

[0183] Step 520: Substitute the second sales volume into the Poisson distribution function, and calculate the remaining quantity of the goods in the warehouse that are about to expire one day before the saleable date based on the substituted Poisson distribution function.

[0184] Optionally, the expiring merchandise may be merchandise that is expiring the next day.

[0185] There may be multiple situations in step 520, such as not considering the loss of the original end-of-period inventory, or considering the loss of the original end-of-period inventory, etc., which are not limited here.

[0186] The following describes how to calculate the quantity of remaining near-expiry goods under two possible scenarios.

[0187] (a) Without considering the loss of the original ending inventory, the estimated number of near-expired goods remaining in the warehouse, near_expired_num_1, is calculated as follows:

[0188]

[0189] Where P(x=k) is the probability of selling k items; s -k) and is the number of expiring goods remaining after selling k items; since the beginning inventory can be a decimal, from the above analysis, with a probability of ω, the beginning inventory is s (s rounded up), and with a probability of 1-ω, the beginning inventory is s (s is rounded down). Therefore, with a probability of 1-ω, the number of remaining expiring goods is With a probability of ω, the number of remaining expiring goods is

[0190] The above method calculates the estimated number of near-expired goods remaining in the warehouse (near_expired_num_1) without considering the loss of the original end-of-period inventory. However, if the possible loss of the original end-of-period inventory is not considered, the calculated near-expiry rate for goods with a short sale period will be too high. Therefore, it is necessary to consider the loss of the original end-of-period inventory, which is case (b).

[0191] (b) Taking into account the loss of the original ending inventory, the estimated number of near-expired goods remaining in the warehouse, near_expired_num, is calculated as follows:

[0192] Step 521: Calculate near_expired_num_1 using the same method as in the above case (a).

[0193] Step 522: Calculate the expected number of lost components, loss_num_1.

[0194] Since the latest loss occurs after the product's sellable period - 1 day, the expected number of lost items, loss_num_1, is calculated as follows:

[0195]

[0196] Where P(x=k) is the probability of selling k items; A -k) and is the number of expiring goods lost after selling k items; since the beginning inventory can be a decimal, from the above analysis, we can see that under the probability of ω', the beginning inventory is (A rounded up), with a probability of 1-ω′, the beginning inventory is A (A is rounded down). Therefore, with a probability of 1-ω′, the number of lost expiring goods is Under the probability of ω', the number of lost expiring goods is

[0197] Step 523: Calculate the quantity of remaining expiring merchandise in the warehouse.

[0198] The quantity of remaining expiring goods in the warehouse is calculated using the following formula:

[0199] near_expired_num=near_expired_num_1-loss_num_1

[0200] Among them, near_expired_num is the number of near-expired goods remaining in the warehouse; near_expired_num_1 is the number of near-expired goods remaining in the warehouse without considering losses; loss_num_1 is the number of near-expired goods that have been lost.

[0201] Step 530: Determine the near-expiry rate based on the remaining quantity of near-expiry merchandise and the beginning inventory.

[0202] There are many ways to determine the near-expiry rate in step 530. For example, the cost and price of the product may be considered, or the cost and price of the product may not be considered, etc., and there is no limitation here.

[0203] The following describes how to calculate the near-term rate under two possible circumstances.

[0204] (a) Without considering the cost and price of the goods, the expiry rate is calculated as follows:

[0205]

[0206] Right now:

[0207] Considering the loss of the original ending inventory in step 520,

[0208] P2 = near_expired_num / s

[0209] Among them, P2 is the rate of premature ejaculation;

[0210] Alternatively, in step 520, without considering the loss of the original ending inventory,

[0211] P2 = near_expired_num_1 / s

[0212] (b) Taking into account the cost and price of the goods, the expiry rate is calculated as follows:

[0213] Step 531: Obtain the cost and price of the product.

[0214] Step 532: Calculate the total cost of the expiring commodities based on the quantity of the expiring commodities and the cost of the commodities.

[0215] Considering the loss of the original ending inventory in step 520,

[0216] C = near_expired_num*cost, where cost is the cost of the product and C is the total cost.

[0217] Alternatively, in step 520, without considering the loss of the original ending inventory,

[0218] C = near_expired_num_1*cost

[0219] Step 533: Calculate the total price of the beginning inventory based on the beginning inventory and the price of the goods.

[0220] Optionally, P = s*price, where price is the product price and P is the total price.

[0221] Step 534: Calculate the near-expiry rate based on the total cost of the near-expiry goods and the total price of the beginning inventory.

[0222] Considering the loss of the original ending inventory in step 520,

[0223] P2=(near_expired_num*cost) / (s*price)

[0224] Alternatively, in step 520, without considering the loss of the original ending inventory,

[0225] P2=(near_expired_num_1*cost) / (s*price)

[0226] Calculation method 2: The quantity distribution of goods follows Gaussian distribution N(μ,σ 2 ) in the case of the calculation of the near-expiry rate.

[0227] Here, the commodity quantity distribution obeys the Gaussian distribution function, which means that the commodity quantity is greater than the first threshold. In other words, the commodity can be considered to be a high-selling commodity.

[0228] Figure 6 The quantity distribution of goods follows Gaussian distribution N(μ,σ 2 ), the flowchart of the calculation of the near-term rate is as follows:

[0229] Step 610: Predict the second sales volume between the current time and the day before the latest available sale date of the product;

[0230] The average remaining number of units between the current time and the day before the latest saleable date of the product can be predicted using accumulation, multiplication, power function, exponential function, etc., without any restrictions here.

[0231] Optionally, the first sales volume is calculated using the following formula:

[0232]

[0233] Among them, sp is the saleable period of the product, that is, the maximum time during which the sale can occur. For example, some aquatic products are only sold on the same day, so sp = 1. For example, some root vegetables have a saleable period of sp = 3, which means that the product can be sold for a total of 3 days after it arrives at the forward warehouse.

[0234] Step 620: Predict the fluctuation between the quantity of goods sold and the demand for the goods between the current time and the day before the latest available sale date of the goods;

[0235] Current methods for calculating inventory indicators often ignore the volatility of product demand, such as seasonality, market trends, and changes in consumer behavior. This results in an inability to accurately estimate inventory risk. Therefore, this embodiment considers the fluctuation between the quantity of goods sold and the demand for the goods, resulting in more accurate calculation results.

[0236] Optionally, the fluctuation between the quantity of goods sold and the quantity demanded can be calculated using the standard deviation.

[0237] The calculation method is as follows:

[0238]

[0239] Where σ1 is the standard deviation of demand.

[0240] Step 630: Substitute the second sales volume and the fluctuation value into the Gaussian distribution function, and calculate the quantity of the expiring merchandise remaining in the warehouse one day before the saleable date of the merchandise in the warehouse based on the substituted Gaussian distribution function.

[0241] There may be multiple situations in step 630, such as not considering the loss of the original end-of-period inventory, or considering the loss of the original end-of-period inventory, etc., which are not limited here.

[0242] The following describes how to calculate the quantity of remaining near-expiry goods under two possible scenarios.

[0243] (a) Without considering the loss of the original ending inventory, the estimated number of near-expired goods remaining in the warehouse, near_expired_num_1, is calculated as follows:

[0244]

[0245] The above method calculates the estimated number of near-expired goods remaining in the warehouse (near_expired_num_1) without considering the loss of the original end-of-period inventory. However, if the possible loss of the original end-of-period inventory is not considered, the calculated near-expiry rate for goods with a short sale period will be too high. Therefore, it is necessary to consider the loss of the original end-of-period inventory, which is case (b).

[0246] (b) Taking into account the loss of the original ending inventory, the estimated number of near-expired goods remaining in the warehouse, near_expired_num, is calculated as follows:

[0247] Step 631: Calculate near_expired_num_1 using the same method as in the above case (a).

[0248] Step 632: Calculate the expected number of lost components, loss_num_1.

[0249] The expected number of loss components loss_num_1 is calculated as follows:

[0250]

[0251] Step 633: Calculate the number of near-expired products remaining in the warehouse, namely, near_expired_num.

[0252] The quantity of remaining expiring goods in the warehouse is calculated using the following formula:

[0253] near_expired_num=near_expired_num_1-loss_num_1

[0254] Among them, near_expired_num is the number of remaining near-expiring goods in the warehouse; near_expired_num_1 is the number of remaining near-expiring goods in the warehouse without loss; loss_num_1 is the number of near-expiring goods with loss.

[0255] Step 640: Determine the expiring rate based on the remaining quantity of expiring merchandise and the beginning inventory.

[0256] There are many ways to determine the near-expiry rate in step 640. For example, the cost and price of the product can be considered, or the cost and price of the product can be not considered, etc., and there is no limitation here.

[0257] The following describes how to calculate the near-term rate under two possible circumstances.

[0258] (a) Without considering the cost and price of the goods, the expiry rate is calculated as follows:

[0259]

[0260] Right now:

[0261] Considering the loss of the original ending inventory in step 630,

[0262] P2 = near_expired_num / s

[0263] Among them, P2 is the rate of premature ejaculation;

[0264] Alternatively, in step 630, without considering the loss of the original ending inventory,

[0265] P2 = near_expired_num_1 / s

[0266] (b) Taking into account the cost and price of the goods, the expiry rate is calculated as follows:

[0267] Step 641: Obtain the cost and price of the product.

[0268] Step 642: Calculate the total cost of the expiring goods based on the quantity of the expiring goods and the cost of the goods.

[0269] Considering the loss of the original ending inventory in step 630,

[0270] C = near_expired_num*cost, where cost is the cost of the product and C is the total cost.

[0271] Alternatively, in step 630, without considering the loss of the original ending inventory,

[0272] C = near_expired_num_1*cost

[0273] Step 643: Calculate the total price of the beginning inventory based on the beginning inventory and the price of the goods.

[0274] Optionally, P = s*price, where price is the product price and P is the total price.

[0275] Step 644: Calculate the near-expiry rate based on the total cost of the near-expiry goods and the total price of the beginning inventory.

[0276] Considering the loss of the original ending inventory in step 630,

[0277] P2=(near_expired_num*cost) / (s*price)

[0278] Alternatively, in step 630, without considering the loss of the original ending inventory,

[0279] P2=(near_expired_num_1*cost) / (s*price)

[0280] Loss rate without intervention

[0281] There are multiple ways to calculate the loss rate without intervention, including but not limited to the following calculation methods 1 and 2.

[0282] Calculation method 1: Calculation of loss rate without intervention when the quantity distribution of goods follows the Poisson distribution P(λ).

[0283] Here, the commodity quantity distribution obeys the Poisson distribution function, which means that the commodity quantity is less than or equal to the first threshold. In other words, the commodity can be considered as a medium-to-low-selling commodity.

[0284] Figure 7 This is a flowchart for calculating the loss rate without intervention when the quantity distribution of goods follows the Poisson distribution P(λ). The specific steps are as follows:

[0285] Step 710: Predicting the third sales volume between the current time and the product's saleable period;

[0286] The average number of remaining units between the current moment and the product's saleable period can be predicted using calculation methods such as accumulation, multiplication, power function, exponential function, etc., without any restrictions here.

[0287] Optionally, the first sales volume is calculated using the following formula:

[0288]

[0289] Here, sp is the saleable period of the product. For example, a product that arrives at the store at 8:00 on the 7th can be sold on the 7th, 8th, and 9th. If it is still not sold by the closing time on the evening of the 9th, it will be lost (at this time, the product has been tried to be sold for a total of 3 days, sp = 3). If it has been sold on the 7th, no loss will occur.

[0290] Step 720: Substitute the third sales volume into the Poisson distribution function, and calculate the loss quantity of goods in the warehouse based on the substituted Poisson distribution function.

[0291] There may be multiple situations in step 720, such as not considering the loss of the original end-of-period inventory, or considering the loss of the original end-of-period inventory, etc., which are not limited here.

[0292] The following describes how to calculate the loss quantity of goods in two possible situations.

[0293] (a) Without considering the loss of the original ending inventory, the estimated loss quantity loss_num_2 is calculated as follows:

[0294]

[0295] Where P(x=k) is the probability of selling k items; s -k) and is the number of expiring goods remaining after selling k items; since the beginning inventory can be a decimal, from the above analysis, we can see that under the probability of ω, the beginning inventory is (s is rounded up), with a probability of 1-ω, the beginning inventory is s (s is rounded down). Therefore, with a probability of 1-ω, the number of remaining expiring goods is With a probability of ω, the number of remaining expiring goods is

[0296] The above method calculates the estimated loss quantity loss_num_2 for each item without considering the loss of the original ending inventory. However, if the possible loss of the original ending inventory is not considered, the loss rate calculated without intervention will be too high for items with a short saleable period. Therefore, it is necessary to consider the loss of the original ending inventory, which is case (b).

[0297] (b) Taking into account the loss of the original ending inventory, the estimated loss quantity loss_num is calculated as follows:

[0298] Step 721: Calculate loss_num_2 using the same method as in the above case (a).

[0299] Step 722: Calculate the expected number of lost components, loss_num_1.

[0300] The calculation method of loss_num_1 is the same as the calculation method of loss_num_1 in step 522.

[0301] End-of-period inventory loss refers to the difference between the actual inventory quantity and the book inventory quantity due to various reasons (such as loss or theft). Since loss occurs at most one day after the sellable period of the product, the calculation of λ in loss_num_1 is the same as in step 510, and the value of loss_num_1 is also the same as the value of loss_num_1 in step 522.

[0302] Step 723: Calculate the loss quantity of goods in the warehouse.

[0303] The loss quantity of goods in the warehouse is calculated by the following formula:

[0304] loss_num=loss_num_2-loss_num_1

[0305] Among them, loss_num is the number of goods lost in the warehouse, loss_num_2 is the number of goods lost in the warehouse without end-of-period inventory loss, and loss_num_1 is the number of goods lost in end-of-period inventory.

[0306] Step 730: Determine the loss rate without intervention based on the loss quantity of the goods and the initial inventory.

[0307] There are many ways to determine the loss rate without intervention in step 730. For example, the cost and price of the goods can be considered, or the cost and price of the goods can be not considered, etc., and there is no limitation here.

[0308] The following describes two possible methods for calculating the loss rate without intervention.

[0309] (a) Without considering the cost and price of the goods, the loss rate without intervention is calculated as follows:

[0310]

[0311] Right now:

[0312] Considering the loss of the original ending inventory in step 720,

[0313] P3=loss_num / s

[0314] Among them, P3 is the loss rate without intervention.

[0315] Alternatively, in step 720, without considering the loss of the original ending inventory,

[0316] P3=loss_num_2 / s

[0317] (b) Taking into account the cost and price of the goods, the loss rate without intervention is calculated as follows:

[0318] Step 731: Obtain the cost and price of the product.

[0319] Step 732: Calculate the total cost of product loss based on the number of damaged products and the cost of the products.

[0320] Considering the loss of the original ending inventory in step 720,

[0321] C = loss_num * cost, where cost is the cost of the product and C is the total cost.

[0322] Alternatively, in step 720, without considering the loss of the original ending inventory,

[0323] C=loss_num_2*cost

[0324] Step 733: Calculate the total price of the beginning inventory based on the beginning inventory and the price of the goods.

[0325] Optionally, P = s*price, where price is the product price and P is the total price.

[0326] Step 734: Calculate the loss rate without intervention based on the total cost of product loss and the total price of the initial inventory.

[0327] Considering the loss of the original ending inventory in step 720,

[0328] P3=(loss_num*cost) / (s*price)

[0329] Alternatively, in step 720, without considering the loss of the original ending inventory,

[0330] P3=(loss_num_2*cost) / (s*price)

[0331] Calculation method 2: The quantity distribution of goods follows Gaussian distribution N(μ,σ 2 ), calculation of loss rate without intervention.

[0332] Here, the commodity quantity distribution obeys the Gaussian distribution function, which means that the commodity quantity is greater than the first threshold. In other words, the commodity can be considered to be a high-selling commodity.

[0333] Figure 8 The quantity distribution of goods follows Gaussian distribution N(μ,σ 2 ), the flowchart of loss rate calculation without intervention is as follows:

[0334] Step 810: Predicting the third sales volume between the current time and the product's saleable period;

[0335] The average number of remaining units between the current moment and the product's saleable period can be predicted using calculation methods such as accumulation, multiplication, power function, exponential function, etc., without any restrictions here.

[0336] Optionally, the first sales volume is calculated using the following formula:

[0337]

[0338] Here, sp is the saleable period of the product. For example, a product that arrives at the store at 8:00 on the 7th can be sold on the 7th, 8th, and 9th. If it is still not sold by the closing time on the evening of the 9th, it will be lost (at this time, the product has been tried to be sold for a total of 3 days, sp = 3). If it has been sold on the 7th, no loss will occur.

[0339] Step 820: Predict the fluctuation between the quantity of goods sold and the demand for the goods between the current time and the period when the goods are available for sale;

[0340] Current methods for calculating inventory indicators often ignore the volatility of product demand, such as seasonality, market trends, and changes in consumer behavior. This results in an inability to accurately estimate inventory risk. Therefore, this embodiment considers the fluctuation between the quantity of goods sold and the demand for the goods, resulting in more accurate calculation results.

[0341] Optionally, the fluctuation between the quantity of goods sold and the quantity demanded can be calculated using the standard deviation.

[0342] The calculation method is as follows:

[0343]

[0344] Where σ1 is the standard deviation of demand.

[0345] Step 830: Substitute the third sales volume and the fluctuation value into the Gaussian distribution function, and calculate the loss quantity of the goods in the warehouse on the saleable date of the goods in the warehouse based on the substituted Gaussian distribution function.

[0346] There may be multiple situations in step 830, such as not considering the loss of the original end-of-period inventory, or considering the loss of the original end-of-period inventory, etc., which are not limited here.

[0347] The following describes how to calculate the loss quantity of goods in two possible situations.

[0348] (a) Without considering the loss of the original ending inventory, the estimated loss quantity loss_num_2 is calculated as follows:

[0349]

[0350] The above method calculates the estimated loss quantity loss_num_2 for each item without considering the loss of the original ending inventory. However, if the possible loss of the original ending inventory is not considered, the loss rate calculated without intervention will be too high for items with a short saleable period. Therefore, it is necessary to consider the loss of the original ending inventory, which is case (b).

[0351] (b) Taking into account the loss of the original ending inventory, the estimated number of expiring goods remaining in the warehouse, loss_num, is calculated as follows:

[0352] Step 831: Calculate loss_num_2 using the same method as in the above case (a).

[0353] Step 832: Calculate the expected number of lost components, loss_num_1.

[0354] The value of loss_num_1 is the same as the value of loss_num_1 in step 632 , and the calculation method is also the same.

[0355] Step 833: Calculate the number loss_num of remaining expiring merchandise in the warehouse.

[0356] The loss quantity of goods in the warehouse is calculated by the following formula:

[0357] loss_num=loss_num_2-loss_num_1

[0358] Among them, loss_num is the number of goods lost in the warehouse, loss_num_2 is the number of goods lost in the warehouse without end-of-period inventory loss, and loss_num_1 is the number of goods lost in end-of-period inventory.

[0359] Step 840: Determine the expiration rate without intervention based on the loss quantity of the product and the initial inventory.

[0360] There are many ways to determine the loss rate without intervention in step 840. For example, the cost and price of the goods can be considered, or the cost and price of the goods can be not considered, etc., and there is no limitation here.

[0361] The following describes two possible methods for calculating the loss rate without intervention.

[0362] (a) Without considering the cost and price of the goods, the loss rate without intervention is calculated as follows:

[0363]

[0364] Right now:

[0365] Considering the loss of the original ending inventory in step 830,

[0366] P3=loss_num / s

[0367] Among them, P3 is the loss rate without intervention.

[0368] Alternatively, in step 830, without considering the loss of the original ending inventory,

[0369] P3=loss_num_2 / s

[0370] (b) Taking into account the cost and price of the goods, the loss rate without intervention is calculated as follows:

[0371] Step 841: Obtain the cost and price of the product.

[0372] Step 842: Calculate the total cost of product loss based on the number of damaged products and the cost of the products.

[0373] Considering the loss of the original ending inventory in step 830,

[0374] C = loss_num * cost, where cost is the cost of the product and C is the total cost.

[0375] Alternatively, in step 830, without considering the loss of the original ending inventory,

[0376] C=loss_num_2*cost

[0377] Step 843: Calculate the total price of the beginning inventory based on the beginning inventory and the price of the goods.

[0378] Optionally, P = s*price, where price is the product price and P is the total price.

[0379] Step 844: Calculate the loss rate without intervention based on the total cost of product loss and the total price of the initial inventory.

[0380] Considering the loss of the original ending inventory in step 830,

[0381] P3=(loss_num*cost) / (s*price)

[0382] Alternatively, in step 830, without considering the loss of the original ending inventory,

[0383] P3=(loss_num_2*cost) / (s*price)

[0384] The following is a specific example to illustrate the calculation of out-of-stock rate, near-expiry rate, and loss rate without intervention.

[0385] Assume that for a product with a 2-day sellable period at a forward warehouse, replenishment can occur daily. The estimated ending inventory of A at a forward warehouse today (T0) is 1.1 units (today's sales have not yet been completed, so the expected future sales volume may be a non-integer). The replenishment quantity of B tomorrow (T1) is 5 units. The expected sales of the next two days (T1 and T2) are 4 and 3.5 units, respectively. The sales ratio (sale_rate) at 16:00 is 70%.

[0386] Due to the low sales volume, we assume that sales volume follows a Poisson distribution. The estimated out-of-stock rate, estimated near-expiry rate, and estimated near-expiry loss rate without intervention are calculated as follows:

[0387] 1. Out-of-stock rate of T1:

[0388] Since replenishment can occur every day, there will be new arrivals in T2 that will affect the 16:00 stock-out situation of T2, so we only focus on the estimated stock-out situation of T1.

[0389] Step 1.1: Calculate the beginning inventory s at T1.

[0390] Beginning inventory of T1

[0391] = Previous day's ending inventory A + today's estimated in-transit inventory B

[0392] =1.1+5=6.1

[0393] Step 1.2: Calculate the beginning inventory as The probability ω.

[0394] ω=s%1=6.1%1=0.1

[0395] Step 1.3: Calculate the first sales volume λ.

[0396] λ=sale_rate*4=0.7*4=2.8

[0397] Step 1.4: Substitute the first sales volume λ into the Poisson distribution P(λ). Based on the substituted Poisson distribution P(λ), calculate the probability of sufficient inventory.

[0398] P1=cdf( s -1)+ω*P(x= s )=cdf(6-1)+0.1*P(x= s )

[0399] Step 1.5: Based on the probability of sufficient inventory, determine the out-of-stock rate.

[0400] P2=1-P1==1-cdf(6-1)-0.1*P(x= s )=1-0.934889686635759-0.1*

[0401] 0.04069969383989624 = 0.06104034398025138 = 6.1%

[0402] 2. T2's terminal rate:

[0403] Since replenishment is done on T1, the replenishment inventory at T1 will become expiring inventory at T2 if it is not consumed.

[0404] Step 2.1: Calculate the second sales volume λ.

[0405]

[0406] Step 2.2: Substitute the second sales volume into the Poisson distribution P(λ). Based on the substitutive Poisson distribution P(λ), calculate the number of next-day expiring products remaining in the warehouse at T1 under the assumption of zero loss.

[0407]

[0408] Step 2.3: Calculate the amount of lost near-expiry merchandise.

[0409] The beginning inventory of T1 without considering replenishment is The probability of ω' = A%1 = 1.1%1 = 0.1

[0410] The number of expiring products lost in T1

[0411]

[0412] Step 2.4: Calculate the number of next-day expiring products remaining in T1.

[0413] near_expired_num=near_expired_num_1-loss_num_1

[0414] =2.2843671836226793-0.02747345833310127≈2.257

[0415] Step 2.5: Determine the expiring rate based on the remaining quantity of expiring merchandise and the beginning inventory

[0416] Near-expired rate = near_expired_num / s = 2.257 / 6.1 = 37.00%

[0417] 3. T2 loss rate without intervention

[0418] Since replenishment is done on T1, if the replenishment quantity of T1 is not consumed in T1 and T2, it will become expiring loss inventory in T2.

[0419] Step 3.1: Calculate the third sales volume λ.

[0420]

[0421] Step 3.2: Substitute the third sales volume into the Poisson distribution P(λ). Based on the substituted Poisson distribution P(λ), calculate the number of goods lost in the warehouse without intervention.

[0422]

[0423] Step 3.3: Calculate the number of lost items.

[0424] The beginning inventory of T1 without considering replenishment is The probability of ω' = A%1 = 1.1%1 = 0.1

[0425] The number of expiring products lost in T1

[0426]

[0427] Step 3.4: The amount of loss of goods in the warehouse.

[0428] loss_num=loss_num_2-loss_num_1=0.496-0.027=0.469

[0429] Step 3.5: Determine the loss rate without intervention based on the number of lost items and the initial inventory.

[0430] Loss rate without intervention = loss_num / s = 0.469 / 6.1 = 0.07688524590163934 = 7.69%

[0431] The above describes the calculation methods for out-of-stock rate, near-expiry rate, and uninterventional shrinkage rate. The following describes the application of different inventory indicators. Specifically, predicting inventory risk for a product based on its values ​​under one or more inventory indicators can be used. For example, this can be used to assess the rationality of inventory parameters such as stocking targets, compare the pros and cons of different replenishment strategies, or guide sales activities.

[0432] Understandably, there are many ways to predict inventory risk for a product, including building statistical models to predict future inventory risk, training models based on historical data to predict inventory risk, assessing potential inventory risk in various scenarios, and predicting potential overstocking risks. The following describes three possible application scenarios for predicting inventory risk for a product.

[0433] Application 1: Use inventory indicators to evaluate warehouse replenishment.

[0434] When developing a replenishment strategy for the forward warehouse's inventory over a specific period, consider short-term estimates of the ending inventory on the day before the target stocking date, the inventory in transit on the stocking date (i.e., the current replenishment volume), and sales volume. By estimating the warehouse's out-of-stock rate, near-expiry rate, and uninterrupted loss rate, and comparing the average of these results over time, you can determine which replenishment strategy is most reasonable and effective.

[0435] To determine the replenishment volume for the warehouse, the short-term estimated ending inventory for the day before the target stocking date, the inventory in transit on the stocking date (i.e., the current replenishment volume), and sales volume need to be evenly distributed to each store. By calculating the difference between each store's historical sales and actual sales, we can derive the standard deviation of demand, and thus calculate the three inventory risk indicators for the warehouse under different replenishment strategies. Ultimately, the optimal replenishment strategy should be selected, which strikes a balance between out-of-stock rates and near-expiry rates.

[0436] Application 2: Use inventory indicators to evaluate inventory parameters.

[0437] Given a series of initial end-of-period inventory levels, average midweek and weekend sales are used as in-transit inventory. Three inventory risk indicators are calculated based on the average midweek and weekend sales. This allows us to establish a relationship between inventory and out-of-stock / near-expiry rates for the product-warehouse dimension. Selecting an interval that meets the target out-of-stock level and target near-expiry level is considered a reasonable inventory parameter range. Furthermore, adjustments to replenishment recommendations are provided to reduce the risk of near-expiry stock-outs.

[0438] First, determine the initial ending inventory, and use the average sales volume on working days (generally Monday to Friday) and rest days (generally Saturday and Sunday) as the in-transit inventory; then, based on these data, calculate the inventory risk indicators. For example, establish the relationship between inventory and out-of-stock rate and near-expiry rate. By selecting an interval that meets the target out-of-stock rate and target near-expiry rate, reasonable inventory parameters can be determined, that is, a certain range of inventory levels can be determined within which the demand for reducing out-of-stock risk can be met while the proportion of near-expiry goods can be controlled. In this way, a balance can be achieved between reducing out-of-stocks and avoiding too many near-expiry goods (reducing waste and losses).

[0439] Application three: Use inventory indicators as activity reference.

[0440] By making short-term forecasts of the day's end-of-period inventory, using the remaining quantity at the end of a batch as in-transit inventory, and estimating the remaining sellable time and sales volume for that batch, we can more accurately predict the near-expiry rate and uninterventional loss rate for that batch in the forward warehouse. This allows us to determine whether to initiate promotional activities in advance to mitigate potential loss risks. For example, if a high near-expiry rate or a high uninterventional near-expiry loss rate is predicted, discounts and other measures can be used to increase sales and reduce losses.

[0441] Based on the same concept, an embodiment of the present application further provides a prediction device, which can execute the prediction method introduced above.

[0442] Please participate Figure 9 , a schematic diagram of the structure of a prediction device provided in an embodiment of the present application is given, such as Figure 9 As shown, the prediction device includes:

[0443] The acquisition module 901 is used to obtain the sales volume of the product.

[0444] Calculation module 902 is used to use a first distribution function to calculate the value of the commodity under a preset indicator if the sales volume of the commodity is less than or equal to a first threshold; and to use a second distribution function to calculate the value of the commodity under the preset indicator if the sales volume of the commodity is greater than the first threshold; wherein the first distribution function is a distribution function that conforms to the probability of a first type of event, the first type of event is an event whose number of occurrences is less than or equal to the first threshold; the second distribution function is a distribution function that conforms to the probability of a second type of event, the second type of event is an event whose number of occurrences is greater than the first threshold; and the preset indicator is an indicator related to the inventory of the commodity.

[0445] Prediction module 903: used to predict the inventory risk of the commodity according to the value of the commodity under the preset indicator.

[0446] In one possible implementation, the calculation module 902 is specifically used to: the preset indicators include one or more of the following: out-of-stock rate indicator, near-expiry rate indicator, loss rate indicator without intervention, the first distribution function is a Poisson distribution function; or, the second distribution function is a Gaussian distribution function.

[0447] In one possible implementation, the calculation module 902 is specifically used to: preset indicators include an out-of-stock rate indicator, and use the first distribution function to calculate the value of the product under the preset indicators, including: predicting the first sales volume between the current moment and the moment of calculating the out-of-stock rate; substituting the first sales volume into the first distribution function, and calculating the value of the first distribution function at each fourth sales volume that is less than the beginning inventory based on the substituted first distribution function; obtaining the probability of sufficient inventory based on the value of the first distribution function at each fourth sales volume that is less than the beginning inventory, and determining the out-of-stock rate based on the probability of sufficient inventory.

[0448] In one possible implementation, the calculation module 902 is specifically used to: preset indicators include an out-of-stock rate indicator, and use a second distribution function to calculate the value of the product under the preset indicators, including: predicting the first sales volume between the current moment and the moment of calculating the out-of-stock rate; predicting the fluctuation value between the number of products sold and the demand for the product between the current moment and the moment of calculating the out-of-stock rate; substituting the first sales volume and the fluctuation value into the second distribution function, and calculating the probability of sufficient inventory based on the substituted second distribution function, and determining the out-of-stock rate based on the probability of sufficient inventory.

[0449] In one possible implementation, the calculation module 902 is specifically used to: predict the first sales volume between the current moment and the moment when the out-of-stock rate is calculated, including: calculating the sales volume before the moment when the out-of-stock rate is calculated and the sales volume for the whole day in a certain historical time period, and calculating the sales volume share; calculating the first sales volume between the current moment and the moment when the out-of-stock rate is calculated based on the daily sales volume before the date when the out-of-stock rate is calculated, and the product of the sales volume and the sales volume share on the date when the out-of-stock rate is calculated.

[0450] In one possible implementation, the calculation module 902 is specifically used to: the preset indicators include an expiring rate indicator, and the first distribution function is used to calculate the value of the product under the preset indicators, including: predicting the second sales volume between the current moment and the day before the latest saleable date of the product, substituting the second sales volume into the first distribution function, and calculating the number of expiring products remaining in the warehouse one day before the saleable date of the products in the warehouse based on the substituted first distribution function, and determining the expiring rate based on the remaining number of expiring products and the beginning inventory.

[0451] In one possible implementation, the calculation module 902 is specifically used for: the preset indicators include an expiring rate indicator, and the second distribution function is used to calculate the value of the product under the preset indicators, including: predicting the second sales volume between the current moment and the day before the latest saleable date of the product, predicting the fluctuation value between the quantity of goods sold between the current moment and the day before the latest saleable date of the product and the demand for the product, substituting the second sales volume and the fluctuation value into the second distribution function, and calculating the number of expiring goods remaining in the warehouse one day before the saleable date of the goods in the warehouse based on the substituted second distribution function, and determining the expiring rate based on the number of remaining expiring goods and the beginning inventory.

[0452] In one possible implementation, the calculation module 902 is specifically used to: determine the near-expiry rate based on the number of remaining near-expiry goods and the beginning inventory, including: determining the ratio of the number of remaining near-expiry goods to the beginning inventory as the near-expiry rate, or obtaining the cost of the goods and the price of the goods; calculating the total cost of the near-expiry goods based on the number of near-expiry goods and the cost of the goods; calculating the total price of the beginning inventory based on the beginning inventory and the price of the goods; and calculating the near-expiry rate based on the total cost of the near-expiry goods and the total price of the beginning inventory.

[0453] In one possible implementation, the calculation module 902 is specifically used for: the preset indicators include a loss rate indicator without intervention, and the first distribution function is used to calculate the value of the product under the preset indicators, including: predicting the third sales volume between the current moment and the saleable period of the product; substituting the third sales volume into the first distribution function, and calculating the loss quantity of the product in the warehouse based on the substituted first distribution function; determining the loss rate without intervention based on the loss quantity of the product and the beginning inventory.

[0454] In one possible implementation, the calculation module 902 is specifically used to: preset indicators include a loss rate indicator without intervention, and use a second distribution function to calculate the value of the product under the preset indicators, including: predicting the third sales volume between the current moment and the saleable period of the product; predicting the fluctuation value between the quantity of goods sold between the current moment and the saleable period of the product and the demand for the product; substituting the third sales volume and the fluctuation value into the second distribution function, and calculating the loss quantity of the goods in the warehouse based on the second distribution function after substitution; determining the loss rate without intervention based on the loss quantity of the goods and the beginning inventory.

[0455] In one possible implementation, calculation module 902 is specifically configured to: determine the no-intervention loss rate based on the number of lost items and the beginning inventory, including: determining the no-intervention loss rate based on the ratio of the number of lost items to the beginning inventory; or obtaining the cost and price of the item; calculating the total cost of lost items based on the number of lost items and the cost of the item; calculating the total price of the beginning inventory based on the beginning inventory and the price of the item; and calculating the no-intervention loss rate based on the total cost of lost items and the total price of the beginning inventory. In this way, taking into account the cost and price of the item makes the calculation result more accurate.

[0456] Optional, beginning inventory as a decimal or integer.

[0457] Please participate Figure 10 , showing a schematic structural diagram of another prediction device provided by an embodiment of the present application, such as Figure 10 As shown, the test device includes: a memory 1001 and a processor 1002, and the processor 1002 is coupled to the memory 1001. The memory 1001 is used to store program instructions, and the processor 1002 is used to call the program instructions stored in the memory 1001 and execute the above test method according to the obtained program.

[0458] Optionally, the prediction device may further include an interface circuit 1003, which may be a transceiver or an input / output interface. The input / output interface is used to input and / or output information, where output can be understood as sending and input can be understood as receiving. The processor 1002 may communicate with other devices in the test device or other devices outside the test device via the interface circuit 1003 to obtain information required for executing the above prediction method.

[0459] When the prediction device 1000 is used to implement Figure 1 When performing the method shown, the processor 1002 is used to implement the functions of the above-mentioned acquisition module 901, calculation module 902, and prediction module 903.

[0460] It is understood that the processor in the embodiments of the present application may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0461] The memory in the embodiments of the present application can be a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an erasable programmable read-only memory, an electrically erasable programmable read-only memory, a register, a hard disk, a mobile hard disk, a compact disc read-only memory (CD-ROM), or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. The storage medium can also be an integral part of the processor.

[0462] Based on the same technical concept, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program or instruction is stored. When the computer program or instruction is executed by a processor, the computer executes the above-mentioned prediction method.

[0463] Based on the same technical concept, an embodiment of the present invention further provides a computer-readable program product, which, when executed, enables a computer to execute the above prediction method.

[0464] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0465] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0466] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0467] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0468] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A prediction method, It is characterized in that The method comprises: Get the sales volume of the product; If the sales volume of the commodity is less than or equal to a first threshold, a first distribution function is used to calculate the value of the commodity under a preset indicator; if the sales volume of the commodity is greater than the first threshold, a second distribution function is used to calculate the value of the commodity under the preset indicator; wherein the first distribution function is a distribution function that the probability of a first type of event conforms to, the first type of event is an event whose number of occurrences is less than or equal to the first threshold, the second distribution function is a distribution function that the probability of a second type of event conforms to, the second type of event is an event whose number of occurrences is greater than the first threshold, and the preset indicator is an indicator related to the inventory of the commodity; The inventory risk of the commodity is predicted according to the value of the commodity under the preset index.

2. The method according to claim 1, It is characterized in that The preset indicators include one or more of the following: out-of-stock rate indicator, near-expiry rate indicator, and loss rate indicator without intervention; The first distribution function is a Poisson distribution function; or, The second distribution function is a Gaussian distribution function.

3. The method according to claim 1, It is characterized in that The preset indicator includes a stock-out rate indicator, and the using of the first distribution function to calculate the value of the commodity under the preset indicator includes: Predict the first sales volume between the current time and the time when the out-of-stock rate is calculated; Substituting the first sales volume into the first distribution function, and calculating the value of the first distribution function at each fourth sales volume that is less than the beginning inventory based on the substituted first distribution function; The probability of sufficient inventory is obtained according to the value of the first distribution function at each fourth sales volume that is less than the initial inventory, and the out-of-stock rate is determined according to the probability of sufficient inventory.

4. The method according to claim 1, It is characterized in that The preset indicator includes a stock-out rate indicator, and the using of the second distribution function to calculate the value of the commodity under the preset indicator includes: Predict the first sales volume between the current time and the time when the out-of-stock rate is calculated; Predict the fluctuation between the quantity of goods sold and the quantity of goods demanded between the current time and the time when the out-of-stock rate is calculated; The first sales volume and the fluctuation value are substituted into the second distribution function, and the probability of sufficient inventory is calculated based on the substituted second distribution function, and the out-of-stock rate is determined based on the probability of sufficient inventory.

5. The method according to claim 3 or 4, It is characterized in that The predicting the first sales volume between the current moment and the moment of calculating the out-of-stock rate includes: Calculate the sales volume before the out-of-stock rate in a certain period of history and the sales volume for the whole day, and calculate the sales volume ratio; The first sales volume between the current moment and the moment of calculating the out-of-stock rate is calculated based on the daily sales volume before the date of calculating the out-of-stock rate and the product of the sales volume on the date of calculating the out-of-stock rate and the sales volume ratio.

6. The method according to claim 1, It is characterized in that The preset index includes an expiration rate index, and the using of the first distribution function to calculate the value of the commodity under the preset index includes: Predict the second sales volume between the current time and the day before the latest available sale date of the product; Substituting the second sales volume into the first distribution function, and calculating the quantity of the expiring commodities remaining in the warehouse one day before the saleable date of the commodities in the warehouse according to the first distribution function after substitution; The near-expiry rate is determined based on the remaining quantity of the near-expiry merchandise and the beginning inventory.

7. The method according to claim 1, It is characterized in that The preset index includes an expiration rate index, and the using of the second distribution function to calculate the value of the commodity under the preset index includes: Predict the second sales volume between the current time and the day before the latest available sale date of the product; Predict the fluctuation between the quantity of goods sold and the demand for goods between the current time and the day before the latest saleable date of the goods; Substituting the second sales volume and the fluctuation value into the second distribution function, and calculating the quantity of the expiring commodities remaining in the warehouse one day before the saleable date of the commodities in the warehouse according to the second distribution function after the substitution; The near-expiry rate is determined based on the remaining quantity of the near-expiry merchandise and the beginning inventory.

8. The method according to claim 6 or 7, It is characterized in that Determining the near-expiry rate according to the remaining quantity of the near-expiry commodities and the initial inventory includes: The ratio of the remaining quantity of the near-expiry commodities to the initial inventory is determined as the near-expiry rate; or, Obtain the cost and price of the goods, calculate the total cost of the expiring goods based on the quantity of the expiring goods and the cost of the goods, calculate the total price of the beginning inventory based on the beginning inventory and the price of the goods, and calculate the expiration rate based on the total cost of the expiring goods and the total price of the beginning inventory.

9. The method according to claim 1, It is characterized in that The inventory index includes a loss rate index without intervention, and the use of the first distribution function to calculate the value of the commodity under the preset index includes: Predict the third sales volume between the current moment and the product’s saleable period; Substituting the third sales volume into the first distribution function, and calculating the loss quantity of the goods in the warehouse according to the first distribution function after substitution; The loss rate without intervention is determined based on the loss quantity of the goods and the initial inventory.

10. The method according to claim 1, It is characterized in that The preset index includes a loss rate index without intervention, and the using of the second distribution function to calculate the value of the commodity under the preset index includes: Predict the third sales volume between the current moment and the product’s saleable period; Predict the fluctuation between the quantity of goods sold and the demand for goods between the current time and the period when the goods are available for sale; Substituting the third sales volume and the fluctuation value into the second distribution function, and calculating the loss quantity of the goods in the warehouse according to the second distribution function after the substitution; The loss rate without intervention is determined based on the loss quantity of the goods and the initial inventory.

11. The method according to claim 9 or 10, It is characterized in that Determining the loss rate without intervention according to the loss quantity and the initial inventory of the commodity includes: Determine the loss rate without intervention based on the ratio of the loss quantity of the commodity to the initial inventory; or, Obtain the cost of the goods and the price of the goods, calculate the total cost of the goods loss according to the loss quantity of the goods and the cost of the goods, calculate the total price of the beginning inventory according to the beginning inventory and the price of the goods, and calculate the loss rate without intervention according to the total cost of the goods loss and the total price of the beginning inventory.

12. A prediction device, It is characterized in that include: The acquisition module is used to obtain the sales volume of the product; a calculation module, configured to use a first distribution function to calculate the value of the commodity under a preset indicator if the sales volume of the commodity is less than or equal to a first threshold, and use a second distribution function to calculate the value of the commodity under the preset indicator if the sales volume of the commodity is greater than the first threshold; wherein the first distribution function is a distribution function that the probability of a first type of event conforms to, the first type of event is an event whose number of occurrences is less than or equal to the first threshold, the second distribution function is a distribution function that the probability of a second type of event conforms to, the second type of event is an event whose number of occurrences is greater than the first threshold, and the preset indicator is an indicator related to the inventory of the commodity; The prediction module is used to predict the inventory risk of the commodity according to the value of the commodity under the preset index.

13. A prediction device, It is characterized in that include: A processor, wherein the processor is coupled to a memory, the memory is used to store a computer program or an instruction, and the processor is used to execute the computer program or the instruction to implement the method according to any one of claims 1 to 11.

14. A computer-readable storage medium, It is characterized in that It stores a computer program executable by a computer device, and when the program is run on the computer device, the computer device executes the steps of any method described in claims 1 to 11.

15. A computer program product, It is characterized in that When the method is executed on a computer, the computer is enabled to execute the method according to any one of claims 1 to 11.

Images