A cigarette specification market status early warning method and model

Abstract

The present invention discloses a method and model for warning the market status of cigarette specifications, and relates to the technical field of market warning control. It includes obtaining a corresponding market status feature set according to the type of cigarette, assigning a weight ratio to each sub-item in the market status feature set according to the cigarette attribute, calculating the share ratio of the number of cigarettes occupied by the seller in the market, and determining the primary threshold according to the weight ratio of each sub-item in the market status feature set and the seller index to obtain the cigarette characteristics. The present invention establishes a comprehensive market status warning model by comprehensively considering the cigarette type, the market status feature set, the share ratio of the seller, and the cigarette characteristic factors, so that the market status can be warned and regulated according to the preset primary threshold and fault tolerance parameters, helping enterprises and decision makers to timely discover potential market risks and take corresponding countermeasures, thereby ensuring the stable development of the cigarette market.

Description

Technical Field

[0001] The present invention relates to the technical field of market early warning control, and in particular to a method and model for early warning of the market status of cigarette specifications. Background Art

[0002] Under the influence of external factors such as economic shocks, the cigarette market usually experiences fluctuations, which will lead to changes in the sales status of cigarette specifications. In order to prevent cigarette brands from performing poorly in the market, it is urgent to develop a cigarette specification market status early warning method and its corresponding model to achieve effective market control.

[0003] After searching, it was found that the Chinese invention patent with publication number "CN117217788A" discloses "a method and system for predicting cigarette demand based on deep learning". The application obtains a customer demand data set by collecting historical ordering data of all customers and all product specifications; normalizes the customer demand data set and divides it into a training set and a test set in proportion; constructs samples based on the training set demand data, uses the sample data to construct a model, and extracts the time series data features of cigarette demand; integrates the customer's basic information, external business information and holiday information, and divides them into different features, and then splices each feature; trains the processed training set data through a deep neural network model; calculates the previous market status indicators; and dynamically adjusts the predicted sales data based on the previous market status indicators.

[0004] In addition, the Chinese invention patent with publication number "CN116739472A" discloses "an inventory warning method, system, computer equipment and medium for warehouse management". This application pre-sets the inventory warning value of each commodity in the target warehouse; receives the query instruction of the target user based on the target warehouse, and displays the current inventory value of the target commodity corresponding to the query instruction through the target table; monitors the current inventory value of the target commodity in real time, and when it is monitored that the current inventory value of the target commodity is less than or equal to its corresponding inventory warning value, issues an inventory warning notification about the target commodity.

[0005] Given that the sales of tobacco products are affected by many factors, including but not limited to tobacco specifications, types, market conditions, dealer inventory ratios and geographical characteristics, however, the currently disclosed early warning and forecasting methods and similar methods focus on factors that are too one-sided in practical applications and do not fully consider the inherent connections between these factors. Therefore, when making market forecasts, it is difficult to fully grasp the complexity and variability of the tobacco market, and it is also impossible to effectively assist companies and decision makers in identifying potential market risks. This leads to significant deviations in the actual early warning process, which in turn affects the sustained and stable development of the tobacco market to a certain extent. Summary of the invention

[0006] The purpose of the present invention is to provide a method and model for early warning of the market status of cigarette specifications, so as to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions: First, a method for early warning of the market status of cigarette specifications is proposed, including: According to the type of cigarettes, obtain the corresponding market status feature set; According to the cigarette attributes, weight ratios are assigned to each sub-item in the market status feature set; Calculate the vendor's share of the number of cigarettes in the market; Determine the primary threshold value based on the weight ratio of each sub-item in the market status feature set and the seller index; Obtaining cigarette characteristics, and determining fault tolerance parameters according to any one or a combination of two of the seller index or the cigarette characteristics; Based on the fault tolerance parameters and primary thresholds, the early warning threshold is calculated to identify abnormal situations in the market status of cigarette specifications; Based on the triggering of the warning threshold, the seller's cigarette market share ratio is adjusted and the cigarette specification level is reconstructed.

[0008] As a further preferred embodiment of the present technical solution, the weight ratio allocation method includes: Establish the thresholds of various cigarette indicators, that is, the time series means corresponding to each sub-item in each market status feature set; According to the difference between the threshold of each cigarette index and the value of each sub-item in the corresponding market status characteristic set, the primary weight ratio is assigned, and the primary weight ratio is arranged in descending order; According to the cigarette attributes, the primary weight ratio of each sub-item in the corresponding market status feature set is integrated; Based on the integration results, the allocation weight ratio of each sub-item in the market status feature set is obtained.

[0009] As a further preferred embodiment of the present technical solution, the weight ratio allocation method includes: According to the characteristics of the time series, each sub-item in the market status feature set is screened; Extract the values ​​of each sub-item in the screening results; Calculate the values ​​of each sub-item in the screening results based on the random forest algorithm and obtain the corresponding auxiliary weight ratio; Combining cigarette attributes with auxiliary weight ratios, determine the allocation weight ratios of each sub-item in the market status feature set.

[0010] As a further preferred embodiment of the present technical solution, the random forest algorithm includes: Set the initial parameters and determine the basic setting values ​​required to build the random forest algorithm; Perform random sampling with replacement on the values ​​of each sub-item to generate multiple sampling sets; Obtain at least one feature from the plurality of sampling sets as a candidate feature, and establish a node of a tree based on the candidate feature; A decision tree is constructed based on the candidate features and the sampling set. During the construction process, the decision tree is continuously split until the values ​​corresponding to the nodes of the tree match the initial parameters. Based on the matched time series nodes, the decision tree nodes are summarized and a random forest is formed; Random forest obtains the final auxiliary weight ratio based on the initial parameter categories.

[0011] As a further preferred embodiment of the present technical solution, the method for determining the primary threshold value includes: Determine the category status of the target cigarette product based on the seller's indicators; According to the type status of the target cigarette product, corresponding sub-items are extracted from the market status feature set and matched; Calculate the initial threshold ratio based on the weight ratio of each sub-item in the matching results and the average of the total market sales of the historical target tobacco products; The calculation formula for the primary threshold is: Primary threshold = (weight ratio of each sub-item × corresponding historical sales total average) / total weight, where the total weight is the sum of the weight ratios of all sub-items in the market status feature set.

[0012] As a further preferred embodiment of the present technical solution, the method for obtaining the fault-tolerant parameters includes: According to the seller's indicators, set sales targets, product category status and expected profit margins for target cigarette products; Use network interconnection technology to obtain the market share of target cigarette products in the areas where the sellers are located, and analyze their fluctuation characteristics at different sales stages; Combine sales targets and historical sales trends during peak and off-seasons to assess the corresponding tolerance parameter ranges; Based on the fault tolerance parameter range and combined with the expected profit amount, determine the fault tolerance parameter.

[0013] As a further preferred embodiment of the present technical solution, the method for obtaining the fault-tolerant parameters includes: Based on the characteristics of cigarettes and based on network interconnection technology, determine the storage conditions of cigarettes in different geographical areas; Construct a correlation diagram between cigarette variation and cigarette storage time based on time series, where the horizontal axis represents storage time and the vertical axis represents variation; Based on market research, collect historical cigarette quality data from a range of vendors; Using historical cigarette quality data and time series data from market research, fit segments are constructed in the correlation graph; Based on the storage conditions, a threshold line segment is constructed in the association graph regarding the cigarette variation and the cigarette storage time; The fault tolerance parameters of cigarettes are obtained based on the fitting line segment and the threshold line segment.

[0014] As a further preferred embodiment of the present technical solution, the calculation method of the warning threshold includes: The early warning threshold is determined by performing a weighted summation of the fault tolerance parameter and the primary threshold; The calculation formula of the warning threshold is expressed as: warning threshold = primary threshold × (1 + fault tolerance parameter); The fault tolerance parameter is determined based on any one or a combination of two of the seller's indicators or cigarette characteristics, and the warning threshold is used to set the abnormal threshold of the market status of cigarette specifications.

[0015] In the second aspect, in order to improve the content proposed in the first aspect, the present invention further proposes a cigarette specification market status early warning model, wherein the cigarette specification market status early warning model uses the cigarette specification market status early warning method disclosed in the first aspect, and includes: The data collection and preprocessing module obtains the corresponding market status feature set according to the type of cigarettes; The weight allocation module allocates weight proportions to each sub-item in the market status feature set according to the cigarette attributes; The overall planning module calculates the proportion of the cigarette quantity that the seller holds in the market; The threshold determination module determines the primary threshold according to the weight ratio of each sub-item in the market status feature set and the seller index; A fault tolerance parameter building module determines the fault tolerance parameter according to any one or a combination of two of the seller index or the cigarette characteristics; The early warning indicator establishment module calculates the early warning threshold based on the fault tolerance parameters and the primary threshold, which is used to identify abnormal situations in the market status of cigarette specifications; The adjustment module adjusts the cigarette market share ratio of the seller based on the triggering of the warning threshold and establishes the cigarette specification grade.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The cigarette specification market status early warning method and model, by comprehensively considering cigarette types, market status feature sets, vendor share ratios and cigarette characteristic factors, establishes a comprehensive market status early warning model, which enables early warning and regulation of market status based on preset primary thresholds and fault tolerance parameters, helping enterprises and decision makers to timely discover potential market risks and take corresponding countermeasures, thus ensuring the stable development of the cigarette market; In addition, the accuracy and robustness of the early warning model are improved by introducing the random forest algorithm to allocate weight ratios. The random forest algorithm constructs decision trees for multiple sampling sets, effectively avoiding the overfitting problem that may be caused by a single decision tree, making the allocation of weight ratios more scientific and reasonable. At the same time, the warning threshold calculation method proposed in the present invention comprehensively considers the influence of fault tolerance parameters and primary thresholds, making the setting of warning thresholds more flexible and adaptable. Through dynamic adjustment of the warning threshold, the present invention can respond to changes in market conditions in a timely manner and provide decision makers with more accurate market warning information. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flowchart of the steps of the method disclosed in the present invention; Figure 2 This is a step diagram of a first implementation mode of step S200 of the present invention; Figure 3 FIG. 2 is a step diagram of a second embodiment of step S200 of the present invention. Figure 4 is an auxiliary illustration of step S230 of the present invention; Figure 5 is an auxiliary illustration of step S560 of the present invention; Figure 6 This is a module composition diagram of the model disclosed in the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Before understanding the technical solution proposed in the present invention, it should be clear that, in the setting of the present invention, cigarette product planning is divided into five different levels, namely "hot, tight, flat, loose, and soft". Among them, the "hot" level of cigarette specifications indicates that the market demand for it is extremely strong, and there is a situation of supply exceeding demand. The "tight" level means that the market supply is slightly tight, but there is still a certain amount of inventory. The "flat" level means that the supply and demand in the market are in a relatively balanced situation, and the inventory quantity is also relatively moderate. The "loose" level means that the supply of this cigarette specification in the market is already in excess, and there is a large backlog of inventory, while the "soft" level means that the market demand for it is very weak, and the sales situation is not ideal.

[0020] like Figure 1 It can be seen that the present invention provides a technical solution: a method for early warning of the market status of cigarette specifications, comprising: steps S100 to S700.

[0021] Step S100: Obtain a corresponding market status feature set according to the type of cigarette.

[0022] It should be clear that in step S100, the types of cigarettes include cigarette brands, cigarette sizes and types of raw materials of cigarettes. In addition, it should be supplemented that in the present invention, step S100 is used to obtain market status characteristic data corresponding to the types of cigarettes. The market status characteristic data reflects the performance of the current cigarette specifications in the market, including customer single product specification ordering demand, cigarette single product specification sales volume, cigarette single product specification sales cost, customer actual retail price, customer price index, customer single product specification gross profit margin, among which customer single product specification ordering demand, cigarette single product specification sales volume, and cigarette single product specification sales cost are from the marketing data of each seller; customer actual retail price, customer price index, and customer single product specification gross profit margin are data from sampled digital stores. By combining the data of the two, the market status characteristics will be better extracted.

[0023] It should be further explained that, in the present invention, the market status feature set is obtained by accessing the cigarette sales database within the Tobacco Bureau and the marketing databases of major retailers, and by applying existing data mining technology to extract and integrate the market status feature data.

[0024] Step S200: assigning weight ratios to the sub-items in the market status feature set according to the cigarette attributes.

[0025] It should be clearly pointed out that in step S200 of the present invention, cigarette attributes include cigarette origin, cigarette price, cigarette flavor preference, economic environment and consumer preference factors. These cigarette attributes serve as the basis for weight allocation and can reveal the competitiveness of different cigarette specifications in the market and the differences in consumer preference. For example, the origin of cigarettes may affect consumers' regional preferences, cigarette prices are directly related to consumers' purchasing power, and cigarette flavor preference is one of the factors that consumers pay the most attention to when choosing cigarette specifications.

[0026] As a preferred embodiment, refer to Figure 2 It can be seen that the method for allocating the weight ratio in step S200 includes: step S201-step S204.

[0027] Step S201: establishing thresholds for various cigarette indicators, that is, the time series means corresponding to each sub-item in each market status feature set.

[0028] It should be clear that the mean of the time series is set to three months (one quarter). Therefore, the mean of the time series corresponding to each sub-item is the sum of the three-month values ​​of each sub-item divided by the three-month value.

[0029] Step S202: assigning primary weight ratios according to the index thresholds of various cigarettes and the numerical differences of the sub-items in the corresponding market status feature set, and the primary weight ratios are arranged in descending order.

[0030] It should be clear that in step S202, the calculation of the primary weight ratio is based on sorting the positive numerical values ​​of the difference values, and assigning corresponding weights in turn.

[0031] Step S203: Integrate the primary weight ratios of the sub-items in the corresponding market status feature set according to the cigarette attributes.

[0032] It should be clear that step S203 is used to classify and arrange the sub-items in the market status feature set according to the cigarette attributes, and obtain the primary weight ratio of the sub-items matching the cigarette attributes as the integration result.

[0033] For example; when the cigarette attribute corresponds to the cigarette flavor preference, the primary weight ratio of the customer single product specification ordering demand and the cigarette single product specification sales volume of the sub-items in the corresponding market status feature set is higher, because the taste preference directly affects the consumer's purchasing decision; when the cigarette attribute corresponds to the economic environment, the primary weight ratio of the customer price index and the cigarette single product specification sales cost of the sub-items in the corresponding market status feature set is more prominent, because changes in the economic environment may directly affect consumers' purchasing power and the sales cost of cigarettes.

[0034] Step S204: Based on the integration result, obtain the allocation weight ratio of each sub-item in the market status feature set.

[0035] As a preferred implementation method, refer to Figure 3 It can be seen that the method for allocating the weight ratio in step S200 also includes: step S210 to step S240.

[0036] Step S210: Filter each sub-item in the market status feature set according to the characteristics of the time series.

[0037] It should be clear that the characteristics of the time series corresponding to step S210 include: holidays. During long holidays, the demand for cigarettes surges, while it is relatively stable on weekdays. By analyzing the characteristics of the time series, those sub-items that are most affected by holidays are screened out from the market status feature set.

[0038] Step S220: extracting the value of each sub-item in the screening result.

[0039] Step S230: Calculate the values ​​of each sub-item in the screening result based on the random forest algorithm, and obtain the corresponding auxiliary weight ratio.

[0040] It should be clear that the random forest in step S230 is an integrated learning method, which is a "forest" formed by combining many decision trees. The random forest constructs multiple decision trees and integrates the prediction results of the decision trees to improve the accuracy and stability of the model.

[0041] Specifically, the method for constructing the random forest in step S230 includes: steps S231 to S236.

[0042] Step S231: Set initial parameters and determine the basic setting values ​​required to construct the random forest algorithm.

[0043] It should be clear that the initial parameters in step S231 include the number of decision trees in the random forest, the maximum depth of the decision tree, and the minimum number of samples required to split a node, wherein the setting of the initial parameters affects the performance of the random forest model and the accuracy of the prediction results.

[0044] Step S232: Perform random sampling with replacement on the values ​​of each sub-item to generate multiple sampling sets.

[0045] Step S233: obtaining at least one feature in the plurality of sampling sets as a candidate feature, and establishing a node of a tree based on the candidate feature.

[0046] Step S234: construct a decision tree based on the candidate features and the sampling set. During the construction process, the decision tree is continuously split until the values ​​corresponding to the nodes of the tree match the initial parameters.

[0047] Step S235: Based on the matched time series nodes, summarize the decision tree nodes and form a random forest.

[0048] Step S236: Random Forest obtains the final auxiliary weight ratio based on the initial parameter category.

[0049] It should be clear that in step S236, the final auxiliary weight ratio for the random forest based on the initial parameter category is obtained by voting or averaging the prediction results of each decision tree through the random forest model to determine the auxiliary weight ratio of each sub-item value.

[0050] As a supplement to step S230, it should be noted that the present invention uses a random forest algorithm to perform weight calculation and auxiliary weight ratio determination. It should be noted that a random forest may contain multiple decision trees, and the final classification result is obtained by the results of multiple decision trees according to the principle of minority obeys majority. For example, there are 100 decision trees in a random forest, 80 of which believe that the status of product specification A in company a is "loose", 10 believe that it is "soft", 5 believe that it is "flat", and 5 believe that it is "tight", then the final status result is "loose".

[0051] In addition, it should be clear that when the random forest algorithm is used to calculate the auxiliary weight ratio, the calculation result is as follows Figure 4 It can be seen that the customer's single product gross profit margin and the customer price index are the two auxiliary weight ratios with the largest weights.

[0052] Step S240: Determine the allocation weight ratio of each sub-item in the market status feature set by combining the cigarette attributes and the auxiliary weight ratio.

[0053] It should be noted that the method for combining cigarette attributes and auxiliary weight ratios in step S240 is the same as the method disclosed in step S203.

[0054] It should be added that, in the present invention, the weight ratio allocation method disclosed in steps S201 to S204 focuses on integrating and obtaining the allocation weight ratio based on the numerical differences between the threshold values ​​of various cigarette indicators and the sub-items in the market status feature set, combined with the cigarette attributes, and the application scenario is conventional cigarette market analysis, while the weight ratio allocation method disclosed in steps S210 to S240 focuses on screening out the sub-items that are more affected by holidays through the analysis of time series characteristics, and using the random forest algorithm to calculate the auxiliary weight ratio, and further combining the cigarette attributes to determine the allocation weight ratio. The application scenario is more complex and is suitable for specific scenarios such as holiday cigarette market demand forecasting.

[0055] Step S300: Calculate the vendor's share of cigarette quantity in the market.

[0056] It should be noted that step S300 in the present invention collects cigarette sales data of each vendor, including sales quantity and sales amount information, and then calculates the cigarette quantity share ratio of each vendor in the market.

[0057] Step S400: Determine a primary threshold value according to the weight ratio of each sub-item in the market status feature set and the vendor index.

[0058] It should be clear that step S400 in the present invention is used to calculate the primary threshold by weighted average based on the weight ratio of each sub-item and seller indicators, such as sales growth rate, customer satisfaction, etc. The primary threshold reflects a benchmark level of the market status of cigarette specifications of each seller under normal circumstances, which is used for subsequent comparison with the early warning threshold to determine whether the market status is abnormal.

[0059] It should be noted that the method for determining the primary threshold in step S400 includes: step S401 to step S403.

[0060] Step S401: Determine the category status of the target cigarette product according to the seller index.

[0061] Step S402: extracting corresponding sub-items from the market status feature set according to the type status of the target cigarette product and performing matching.

[0062] Step S403: Calculate a preliminary threshold ratio based on the weight ratio of each sub-item in the matching result and the average value of the total market sales volume of the historical target tobacco product.

[0063] It should be added that the calculation formula of the primary threshold in step S403 is: primary threshold = (weight ratio of each sub-item × corresponding historical sales total average) / total weight, where the total weight is the sum of the weight ratios of all sub-items in the market status feature set.

[0064] It should be further explained that when step S403 is put into actual operation, for example, in a certain quarter, the category status of the target cigarette product is "popular", then the two sub-items of customer single product specification order demand and cigarette single product specification sales volume are extracted from the market status feature set for matching. Assuming that the weight ratios of these two sub-items are 0.6 and 0.4 respectively, and the corresponding historical sales volume averages are 10,000 boxes and 8,000 boxes respectively, the calculation process of the primary threshold is: primary threshold = (0.6 × 10,000 +0.4 × 8,000) / (0.6 + 0.4) = 9,200 boxes.

[0065] Step S500: Obtain cigarette characteristics, and determine fault tolerance parameters according to any one or a combination of two of the vendor index or the cigarette characteristics.

[0066] It should be noted that in step S500, the cigarette characteristic is the storage condition of the cigarette.

[0067] As a preferred implementation, the method of using the vendor index to determine the fault tolerance parameter in step S500 includes: step S501 to step S504.

[0068] Step S501: According to the indicators of the seller, the sales target, product category status and expected profit margin of the target cigarette product are set.

[0069] Step S502: using network interconnection technology to obtain the market share of the target cigarette product in the area where the seller is located, and analyzing its fluctuation characteristics in different sales stages.

[0070] Step S503: Evaluate the corresponding fault tolerance parameter range based on the sales target and the historical sales trends of the peak and off-seasons.

[0071] Step S504: Based on the fault tolerance parameter range and in combination with the expected profit amount, determine the fault tolerance parameter.

[0072] It should be noted that in step S503, the method for evaluating the fault tolerance parameter interval in combination with the sales target and the trends of the historical sales peak season and off-season is to identify the rules of the sales peak season and off-season through statistics and analysis of historical sales data, and then adjust the range of the fault tolerance parameter interval according to the sales target. For example, if the sales target is high and it is currently in the peak sales season, the fault tolerance parameter interval can be appropriately relaxed to cope with the uncertainty of market demand. Conversely, if the sales target is low and it is currently in the off-season, the fault tolerance parameter interval needs to be tightened to ensure the achievement of the sales target. Through this method, the determination of the fault tolerance parameters can be made more scientific and reasonable, thereby improving the accuracy and reliability of the market status warning of cigarette specifications.

[0073] As a preferred implementation, the method for determining the fault tolerance parameter using cigarette characteristics in step S500 includes: steps S510 to S560.

[0074] Step S510: Determine the storage conditions of cigarettes in different geographical areas based on the characteristics of cigarettes and network interconnection technology.

[0075] Step S520: constructing a correlation diagram between cigarette variation and cigarette storage time based on time series, wherein the horizontal axis represents storage time and the vertical axis represents variation, wherein the variation is mainly reflected by obtaining the tobacco moisture content or mold content in the cigarette.

[0076] Step S530: Based on market research, collect historical cigarette quality data of a series of vendors.

[0077] Step S540: construct a fitting line segment in the correlation graph using the historical cigarette quality data and the time series data of the market research.

[0078] Step S550: constructing a threshold line segment about cigarette variation and cigarette storage time in the association graph based on the storage condition acquisition.

[0079] Step S560: Obtaining the fault tolerance parameter of the cigarette according to the fitting line segment and the threshold line segment.

[0080] It should be clear that reference Figure 5 It can be seen that the fitting line segment and the threshold line segment can reflect the quality change trend and quality threshold range of cigarettes under different storage conditions. Among them, the fitting line segment is constructed based on historical cigarette quality data and time series data, representing the variation of cigarettes under normal conditions as the storage time increases, while the threshold line segment is set based on the storage conditions, representing the maximum variation range allowed for cigarettes under different storage conditions. By comparing the fitting line segment and the threshold line segment, it is intuitively judged whether the current quality status of the cigarette is within the normal range, and then determine whether it is necessary to adjust the storage conditions or take other measures to maintain the quality of cigarettes. This method can comprehensively consider the characteristics and storage conditions of cigarettes, and improve the accuracy and pertinence of the market status warning of cigarette specifications.

[0081] As a preferred implementation, the method of using the vendor index and cigarette characteristics to determine the error tolerance parameter in step S500 is to combine the error tolerance parameters obtained in steps S501 to S504 and steps S510 to S560 (to form a range).

[0082] In addition, it should be noted that the fault tolerance parameters determined by steps S501 to S504 are mainly used in evaluation areas such as cigarette sales targets, product category status, and expected profit margins, with an emphasis on market demand and sales target considerations to ensure that the early warning system can fully take into account the actual conditions and profit targets of sellers, while the fault tolerance parameters determined by steps S510 to S560 are mainly used in the evaluation of cigarette storage conditions, with an emphasis on the physical properties of cigarettes to ensure that cigarettes maintain quality stability during storage. By combining the fault tolerance parameters of these two aspects, a more comprehensive and scientific fault tolerance parameter range can be constructed, which takes into account both market demand and sales targets as well as the physical properties and storage conditions of cigarettes, thereby further improving the accuracy and practicality of cigarette product market status early warnings.

[0083] Step S600: Calculate the warning threshold based on the fault tolerance parameter and the primary threshold to identify abnormal situations in the market status of cigarette specifications.

[0084] It should be clear that the calculation method of the warning threshold in step S600 is to determine the warning threshold by performing weighted summation on the fault tolerance parameter and the primary threshold.

[0085] It should be noted that the calculation formula of the warning threshold is expressed as: warning threshold = primary threshold × (1 + fault tolerance parameter); The fault tolerance parameter is determined based on any one or a combination of two of the seller's indicators or cigarette characteristics, and the warning threshold is used to set the abnormal threshold of the market status of cigarette specifications.

[0086] Specifically, when the warning threshold calculation formula disclosed in step S600 is substituted into actual application, for example, the target cigarette product of a certain seller is specification A, the primary threshold in a certain quarter is 9200 boxes, and the fault tolerance parameter determined according to the seller's indicators is 0.1, then the calculation process of the warning threshold is: warning threshold = 9200 × (1 + 0.1) = 10120 boxes, which means that when the total market sales volume of specification A cigarettes exceeds 10120 boxes, an early warning will be issued to remind the seller to pay attention to changes in market conditions and take timely measures to deal with possible market risks.

[0087] Step S700: Based on the triggering of the warning threshold, the cigarette market share ratio of the seller is adjusted, and the cigarette specification level is reconstructed.

[0088] It should be clear that in the present invention, after the triggering of the warning threshold in step S700, the cigarette supply quantity of the seller in the market is artificially adjusted. For sellers exceeding the warning threshold, their cigarette supply share is appropriately reduced to avoid market overheating or oversupply. For sellers that have not reached the warning threshold, consideration is given to increasing their cigarette supply share to stimulate sales and market share growth. At the same time, it should be added that reconstructing the cigarette specification level in step S700 means re-evaluating and adjusting the status and level of each cigarette specification in the market according to the triggering of the warning threshold. For hot-selling and popular specifications, their level can be upgraded and given more market support and resource allocation. For specifications with poor sales or lackluster market response, consideration can be given to lowering their level and adjusting sales strategies and promotion methods to improve their market competitiveness. Through such adjustments, the market status of cigarette specifications can be made more in line with actual needs and the overall efficiency and benefits of the market can be improved.

[0089] It should be further added that the method of artificially increasing the cigarette market share ratio follows the principle that the total cigarette sales volume remains unchanged, that is, the increase in total sales volume = the decrease in total sales volume.

[0090] Specifically, step one, determine the product specifications to be increased: according to the product specification number, find the product specifications of the same industry and price range as the product specifications to be reduced, and regard the product specifications with a market status of "hot, tight, and flat" as the objects of increase. The increase in each product specification is determined by the sales volume ratio between the adjusted specifications. For example, if it is known that the product specification A to be reduced belongs to the industry A China Tobacco and the price range is 280-400 (inclusive) yuan, and assuming that the reduction is 3,000 pieces, first find the product specifications belonging to Sichuan China Tobacco and the price range is 280-400 (inclusive) yuan among the product specifications with a market status of "hot, tight, and flat". Assuming that there are three product specifications B, C, and D, and the sales ratio of the three is 1:4:1, then product specification B can be increased by 3,000×1 / 6=500 pieces, product specification C can be increased by 2,000 pieces, and product specification D can be increased by 500 pieces. The sales proportion of product specifications = sales of a certain product specification / total sales of qualified product specifications. If there is no product specification in the same industry and price range as the reduced product specification, only one condition is considered, that is, only the same industry or the same price range is met, and two results are obtained.

[0091] Step 2. Determine the branches to be increased: Taking the above C product specification as an example, screen out the branches with a market status of "flat" or above in C product specification, and allocate the initial adjustment value according to the sales share of each branch. For example, the screened companies include 15 of the 17 branches, including a, b, and c. Among them, the sales of branch a accounts for one tenth of the 15 branches, that is, branch a can increase 2000×0.1=200 lines. Similarly, you can calculate how many lines other branches can increase.

[0092] Step 3. Evaluate the status after the increase: increase the sales volume of product C by 5% each time. For example, if the original sales volume of product C of branch company A is 600 cartons, it can be increased by 200 cartons. The increase each time is calculated by 5%, that is, 200×5%=10 cartons, that is, 10 cartons are increased each time. After one increase, the sales volume of company A is 600+10=610 cartons, that is, the new sales volume = original sales volume + the increased amount. Similarly, the regression (customer actual retail price, customer price index, customer single product gross profit margin) model used during the reduction can be used to calculate the new customer single product order demand, cigarette single product sales volume, cigarette single product sales cost, customer actual retail price, customer price index, and customer single product gross profit margin. Substituting the new data of these 6 indicators into the "market status evaluation model" can obtain the market status after the increase.

[0093] Step 4. Adjustment strategy: If the market condition deteriorates (i.e. the condition after adjustment is below "flat"), the adjustment will be stopped. If there is no change, the adjustment will continue.

[0094] The final output results: reduced commodity code, reduced product specification industry name, reduced product specification price range, status after reduction, increased product specification code, affiliated company, original sales volume, adjusted volume, original status, status after increase, increased product specification industry name, and increased product specification price range.

[0095] It should be added that the methods for artificially reducing the cigarette market share ratio include: Step 1: Determine the objects of reduction.

[0096] Step 2. Evaluate the status after the reduction: Assuming that the sales cost of cigarette products remains unchanged, calculate the new customer actual retail price, customer price index, and customer product gross profit margin based on the new cigarette product sales volume and customer product order demand (demand = reduced volume / demand satisfaction rate). There is no direct formula to calculate the customer actual retail price, customer price index, and customer product gross profit margin in the digital store sampling data when sales change at the company level.

[0097] Step 3. Result analysis: If the market status of a certain cigarette specification of a company is improved (from "soft" to "loose" or above), the reduction amount of the specification of the company can be calculated, that is, the total reduction amount = each reduction amount (5% of the initial sales volume) × the number of reductions. If the sales volume drops to 0 and the market status remains unchanged, the status level will be automatically upgraded to a higher level, and so on. The total reduction amount of the reduced specification can be obtained by adding up the reduction amounts of all the branches that have reduced the product.

[0098] In a preferred embodiment, the present invention further proposes a cigarette specification market status early warning model. Figure 6 It is clearly demonstrated in the paper. It should be pointed out that this cigarette specification market status early warning model is based on Figure 1 The disclosed method for early warning of the market status of cigarette specifications is constructed, and it includes the following key modules: The data collection and preprocessing module obtains the corresponding market status feature set according to the type of cigarettes; The weight allocation module allocates weight proportions to each sub-item in the market status feature set according to the cigarette attributes; The overall planning module calculates the proportion of the cigarette quantity that the seller holds in the market; The threshold determination module determines the primary threshold according to the weight ratio of each sub-item in the market status feature set and the seller index; A fault tolerance parameter building module determines the fault tolerance parameter according to any one or a combination of two of the seller index or the cigarette characteristics; The early warning indicator establishment module calculates the early warning threshold based on the fault tolerance parameters and the primary threshold, which is used to identify abnormal situations in the market status of cigarette specifications; The adjustment module adjusts the cigarette market share ratio of the seller based on the triggering of the warning threshold and establishes the cigarette specification grade.

[0099] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is limited by the attached embodiments and their equivalents.

Claims

1. A cigarette specification market status early warning method, characterized in that: include: According to the type of cigarettes, obtain the corresponding market status feature set; According to the cigarette attributes, weight ratios are assigned to each sub-item in the market status feature set; Calculate the vendor's share of the number of cigarettes in the market; Determine the primary threshold value according to the weight ratio of each sub-item in the market status feature set and the seller index; Obtaining cigarette characteristics, and determining fault tolerance parameters according to any one or a combination of two of the vendor index or the cigarette characteristics; Based on the fault tolerance parameters and primary thresholds, the early warning threshold is calculated to identify abnormal situations in the market status of cigarette specifications; Based on the triggering of the warning threshold, the seller's cigarette market share ratio is adjusted and the cigarette specification level is reconstructed.

2. The cigarette specification market status early warning method according to claim 1 is characterized by: The weight ratio allocation methods include: Establish the thresholds of various cigarette indicators, that is, the time series means corresponding to each sub-item in each market status feature set; According to the difference between the threshold of each cigarette index and the value of each sub-item in the corresponding market status characteristic set, the primary weight ratio is assigned, and the primary weight ratio is arranged in descending order; According to the cigarette attributes, the primary weight ratio of each sub-item in the corresponding market status feature set is integrated; Based on the integration results, the allocation weight ratio of each sub-item in the market status feature set is obtained.

3. The cigarette specification market status early warning method according to claim 1 is characterized by: The weight ratio allocation methods include: According to the characteristics of the time series, each sub-item in the market status feature set is screened; Extract the values ​​of each sub-item in the screening results; Calculate the values ​​of each sub-item in the screening results based on the random forest algorithm and obtain the corresponding auxiliary weight ratio; Combining cigarette attributes with auxiliary weight ratios, determine the allocation weight ratios of each sub-item in the market status feature set.

4. The cigarette specification market status early warning method according to claim 3 is characterized by: The random forest algorithm includes: Set the initial parameters and determine the basic setting values ​​required to build the random forest algorithm; Perform random sampling with replacement on the values ​​of each sub-item to generate multiple sampling sets; Obtain at least one feature from the plurality of sampling sets as a candidate feature, and establish a node of a tree based on the candidate feature; A decision tree is constructed based on the candidate features and the sampling set. During the construction process, the decision tree is continuously split until the values ​​corresponding to the nodes of the tree match the initial parameters. Based on the matched time series nodes, the decision tree nodes are summarized and a random forest is formed; Random forest obtains the final auxiliary weight ratio based on the initial parameter categories.

5. The cigarette specification market status early warning method according to claim 1 is characterized by: Methods for determining the primary threshold include: Determine the category status of the target cigarette product based on the seller's indicators; According to the type status of the target cigarette product, corresponding sub-items are extracted from the market status feature set and matched; Calculate the initial threshold ratio based on the weight ratio of each sub-item in the matching results and the average of the total market sales of the historical target tobacco products; The calculation formula for the primary threshold is: Primary threshold = (weight ratio of each sub-item × corresponding historical sales total average) / total weight, where the total weight is the sum of the weight ratios of all sub-items in the market status feature set.

6. The cigarette specification market status early warning method according to claim 1 is characterized by: Methods for obtaining fault-tolerant parameters include: According to the seller's indicators, set sales targets, product category status and expected profit margins for target cigarette products; Use network interconnection technology to obtain the market share of target cigarette products in the areas where the sellers are located, and analyze its fluctuation characteristics at different sales stages; Combine sales targets and historical sales trends during peak and off-seasons to assess the corresponding tolerance parameter ranges; Based on the fault tolerance parameter range and combined with the expected profit amount, the fault tolerance parameters are determined.

7. The cigarette specification market status early warning method according to claim 1 is characterized by: Methods for obtaining fault-tolerant parameters include: Based on the characteristics of cigarettes and based on network interconnection technology, determine the storage conditions of cigarettes in different geographical areas; Construct a correlation diagram between cigarette variation and cigarette storage time based on time series, where the horizontal axis represents storage time and the vertical axis represents variation; Based on market research, collect historical cigarette quality data from a range of vendors; Using historical cigarette quality data and time series data from market research, fit segments are constructed in the correlation graph; Based on the storage conditions, a threshold line segment is constructed in the association graph regarding the cigarette variation and the cigarette storage time; The fault tolerance parameters of cigarettes are obtained based on the fitting line segment and the threshold line segment.

8. The cigarette specification market status early warning method according to claim 1 is characterized by: The calculation method of the warning threshold includes: The early warning threshold is determined by performing a weighted summation of the fault tolerance parameter and the primary threshold; The calculation formula of the warning threshold is expressed as: warning threshold = primary threshold × (1 + fault tolerance parameter); The fault tolerance parameter is determined based on any one or a combination of two of the seller's indicators or cigarette characteristics, and the warning threshold is used to set the abnormal threshold of the market status of cigarette specifications.

9. A cigarette specification market status early warning model, using a cigarette specification market status early warning method according to any one of claims 1 to 8, characterized in that: include: The data collection and preprocessing module obtains the corresponding market status feature set according to the type of cigarettes; The weight allocation module allocates weight proportions to each sub-item in the market status feature set according to the cigarette attributes; The overall planning module calculates the proportion of the cigarette quantity that the seller holds in the market; The threshold determination module determines the primary threshold according to the weight ratio of each sub-item in the market status feature set and the seller index; A fault tolerance parameter building module determines the fault tolerance parameter according to any one or a combination of two of the seller index or the cigarette characteristics; The early warning indicator establishment module calculates the early warning threshold based on the fault tolerance parameters and the primary threshold, which is used to identify abnormal situations in the market status of cigarette specifications; The adjustment module adjusts the cigarette market share ratio of the seller based on the triggering of the warning threshold and establishes the cigarette specification grade.

Images