Take-out platform time-sharing order prediction method and time-sharing order prediction model training method

Abstract

The invention provides a take-out platform time-sharing order prediction method and a time-sharing order prediction model training method, and belongs to the technical field of take-out order quantity prediction.The time-sharing order prediction method comprises the steps that meteorological data and time data of a prediction time period are input into a time-sharing order prediction model, and a take-out order quantity prediction value of the prediction time period is output; and obtaining calibration data, and calibrating the take-out order quantity prediction value in the prediction time period through the calibration data to obtain a final take-out order quantity prediction value in the prediction time period. According to the method, the take-out order quantity of the take-out platform in different periods can be accurately predicted.

Description

Technical Field

[0001] The present invention relates to the technical field of takeaway order quantity prediction, and in particular to a time-sharing order prediction method for a takeaway platform and a time-sharing order prediction model training method. Background Art

[0002] As people's pace of life accelerates, more and more people choose to order food online, which also means a substantial increase in the number of orders, especially during peak meal times such as lunch or dinner, when the frequency of takeout orders is very high. This requires accurately predicting the number of orders that may appear in each period in the future in a certain area in order to equip enough delivery riders to meet the delivery needs of different periods. This is of great significance for solving the problem of takeout order delivery, reducing user waiting time, and improving user satisfaction. Summary of the invention

[0003] The embodiments of the present invention provide a method for predicting time-sharing orders on a food delivery platform and a method for training a time-sharing order prediction model to solve the problem of accurately predicting the number of food delivery orders in each time period.

[0004] In a first aspect, an embodiment of the present invention provides a method for predicting time-sharing orders on a food delivery platform, comprising:

[0005] The meteorological data and time data of the forecast period are input into the time-sharing order forecasting model, and the forecast value of the number of takeout orders in the forecast period is output;

[0006] The calibration data is obtained, and the prediction value of the number of takeout orders for the prediction period is calibrated by the calibration data to obtain the final prediction value of the number of takeout orders for the prediction period.

[0007] In a second aspect, an embodiment of the present invention provides a time-sharing order prediction model training method, comprising:

[0008] Obtain historical takeaway order data from the takeaway platform, divide the 24 hours a day into n preset time periods, and summarize the number of takeaway orders in each preset time period based on the takeaway order data;

[0009] Obtain the meteorological data and time data for each preset time period, perform one-hot encoding on the meteorological data and time data, and standardize the meteorological data and time data using the min-max method so that the result value is mapped between [0-1];

[0010] Calculate the Pearson correlation coefficient between the weather data, time data and the number of takeout orders in each preset time period, retain the weather data and time data whose absolute value of the Pearson correlation coefficient is higher than the Pearson correlation coefficient threshold, and obtain optimized weather data and time data;

[0011] The time-sharing order prediction model is obtained based on the summarized number of takeout orders in each preset time period and the optimized meteorological data and time data training.

[0012] In a third aspect, an embodiment of the present invention provides a terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method described in the first aspect, the second aspect, or any possible implementation of the first aspect or the second aspect are implemented.

[0013] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect, the second aspect, or any possible implementation of the first aspect or the second aspect are implemented.

[0014] The embodiment of the present invention provides a method for predicting time-sharing orders on a food delivery platform and a method for training a time-sharing order prediction model, wherein the meteorological data and time data of the prediction period are input into the time-sharing order prediction model, and the prediction value of the number of food delivery orders in the prediction period is output, thereby predicting the number of food delivery orders in each period. The prediction value of the number of food delivery orders in the prediction period is calibrated using calibration data to obtain the final prediction value of the number of food delivery orders in the prediction period. The prediction value of the number of food delivery orders in the prediction period is calibrated using calibration data, so that the final prediction value of the number of food delivery orders in the prediction period is more accurate, which is conducive to the platform to reasonably allocate the number of delivery riders. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 creative labor.

[0016] Figure 1 It is a flow chart of a method for predicting the number of time-sharing orders on a food delivery platform provided by an embodiment of the present invention;

[0017] Figure 2 It is a flowchart of a method for training a time-sharing order quantity prediction model for a food delivery platform provided by an embodiment of the present invention;

[0018] Figure 3 It is a structural schematic diagram of a device for predicting the number of time-sharing orders on a takeaway platform provided by an embodiment of the present invention;

[0019] Figure 4 It is a structural schematic diagram of a takeaway platform time-sharing order quantity prediction model training device provided by an embodiment of the present invention;

[0020] Figure 5 is a schematic diagram of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0022] In order to make the purpose, technical solutions and advantages of the present invention more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0023] Figure 1 The flowchart of the method for predicting the number of time-sharing orders on a takeaway platform provided by an embodiment of the present invention. Figure 1 As shown, the details are as follows:

[0024] Step 101: input the meteorological data and time data of the forecast period into the time-sharing order forecasting model, and output the forecast value of the number of takeout orders in the forecast period.

[0025] The weather data and time data of the forecast period are input into the time-sharing order forecasting model. Based on the input weather and time data, the expected number of takeout orders in the forecast period is calculated and output, thereby predicting the number of takeout orders in each period of the day. Through the time-sharing order forecasting model, the food delivery platform can more accurately predict the number of takeout orders in different time periods, thereby better planning resources, optimizing delivery routes and dispatching delivery personnel, and improving overall operational efficiency.

[0026] Data analysis that uses meteorological and time factors to predict the volume of takeout orders can make more accurate demand forecasts. The takeout order forecast value can help food delivery platforms or merchants prepare in advance, such as adjusting transportation capacity, optimizing food supply, adjusting marketing strategies, etc., thereby improving service quality and operational efficiency.

[0027] Meteorological data includes sunny, cloudy, light rain, heavy rain, rainstorm, windy, light snow and heavy snow, etc. Under extreme weather conditions, such as typhoons and rainstorms, the volume of takeout orders usually increases by 20% to 30%, and even in special weather conditions such as light rain and light snow, there will be an increase of 5% to 15%.

[0028] The impact of time periods on the number of takeout orders is significant. Consumers' dining scenarios are becoming more and more diverse. In addition to the traditional lunch and dinner peak hours, the number of takeout orders during non-conventional dining hours such as breakfast, afternoon tea, and supper is also increasing. For example, the survey shows that 57.0% of consumers choose to order takeout as a dinner solution when working overtime, which shows that there is also a significant demand for takeout during non-peak hours on weekdays.

[0029] The time-sharing order prediction model is trained by summarizing the number of takeout orders in each preset time period and optimizing the meteorological data and time data.

[0030] Step 102: Acquire calibration data, and calibrate the forecast value of the number of takeout orders for the forecast period using the calibration data to obtain the final forecast value of the number of takeout orders for the forecast period.

[0031] The applicant has found through research that, in addition to the impact of weather and time period factors involved in step 101 on the number of takeout orders, there are other factors that will affect the number of takeout orders, such as social and cultural events, technical problems of the takeout platform, population density, temperature, etc. Social and cultural events such as large-scale sports events, concerts, political events, etc. may temporarily change people's demand for takeout; technical problems of the takeout platform, such as server downtime or application failure, may temporarily affect the number of takeout orders. The impact of social and cultural events, technical problems of the takeout platform, etc. on the number of takeout orders is uncertain. In order to be more accurate, it is more accurate to select population density and temperature to calibrate the number of takeout orders.

[0032] The calibration data is used to calibrate the forecast value of the number of takeout orders for the forecast period to obtain the final forecast value of the number of takeout orders for the forecast period. The calibration data is used to calibrate the forecast value of the number of takeout orders for the forecast period to make the final forecast value of the number of takeout orders for the forecast period more accurate.

[0033] In this embodiment, it can help the food delivery platform better understand the order demand in different time periods, so as to optimize the operation strategy and resource allocation. For example, based on the order volume forecast for the time period, the platform can adjust the riders' work shifts to ensure that there is enough capacity to meet the order demand during peak hours. At the same time, the platform can also adjust the marketing strategy based on the forecast results, such as launching promotional activities during the period with lower order volume to stimulate demand.

[0034] The preferred implementation of each step of this embodiment is described in detail below.

[0035] In one possible implementation, the calibration data includes:

[0036] The ratio of the average number of takeout orders in each historical period to the average number of total daily takeout orders, population density, and temperature.

[0037] Among them, by cooperating with operators such as China Mobile, China Unicom, and China Telecom, we collect usage data of mobile phone users to collect population density, including data on calls, text messages, and data traffic usage. For example, when population density is based on a business district, the data collected on the population density of the business district should include a timestamp to determine the specific activity time of the user in the business district. When population density is based on statistical information on the location distribution of office buildings, the data collected on the population density of office buildings should include a timestamp to determine the specific activity time of the user in the office building area. When population density is based on holiday information, the data collected on holiday population density should include a timestamp to determine the specific activity time and location of the user during the holidays.

[0038] Areas with high population density usually mean more potential customers, which can lead to an increase in the number of food delivery orders. For example, city centers or business districts usually have higher population densities than suburban or rural areas, so food delivery demand may be stronger in these areas.

[0039] The demand for food delivery can vary greatly during different time periods. For example, lunch and dinner time may be peak times for food delivery orders, while late night or early morning may be low times.

[0040] Population density and time periods may affect each other. For example, in areas with high population density, food delivery demand may remain high even during off-peak hours because there are more residents and workers who need food services. In addition, specific time periods (such as holidays) may lead to further increases in food delivery demand in areas with high population density.

[0041] In a possible implementation, calibrating the forecast value of the number of takeout orders in the forecast period by using calibration data includes:

[0042] Based on the impact of changes in population density and temperature on the number of takeout orders, the ratio of the average number of takeout orders in each historical period to the average number of daily takeout orders is used to calibrate the forecast value of the number of takeout orders in the forecast period. The formula is:

[0043] M=y[|aa ′ |Δd a +|bb ′ |Δd b ]M ′

[0044] Where M is the predicted value of the number of takeout orders in the calibrated forecast period, y is the ratio of the average number of takeout orders in the historical forecast period to the average number of total takeout orders per day, a is the average daily temperature on the forecast day, and a ′ is the historical daily average temperature, Δd ais the weight coefficient of the daily average temperature on the takeaway order volume, b is the population density during the forecast period, and b ′ is the average population density during the historical forecast period, Δd b is the weight coefficient of population density on the number of takeout orders, M ′ It is the predicted value of the number of takeout orders for the preset time period.

[0045] Δd a Indicates the degree of influence of the average temperature on the order forecast value. When the average temperature rises or falls, that is, the temperature is higher or lower, people will feel uncomfortable after experiencing different ranges of temperature, and the number of takeaway orders will increase. Δd b Indicates the degree of influence of changes in population density on order forecast values. In areas with high population density, the corresponding demand for takeout may increase, for example, Δd a Can be set to 0.1, Δd b Can be set to 0.2.

[0046] y is the ratio of the average volume of takeout orders in the predicted period in the historical data to the average volume of total takeout orders. The order volumes in different periods of the day will be different. For example, lunch time is the peak period for takeout, so the ratio of the order volume in the noon period to the total order volume of the day will be higher.

[0047] In another possible implementation, calibrating the forecast value of the number of takeout orders in the forecast period by using calibration data includes:

[0048] Based on the impact of changes in population density and temperature on the number of takeout orders, the ratio of the average number of takeout orders in each historical period to the average number of daily takeout orders is used to calibrate the forecast value of the number of takeout orders in the forecast period. The formula is:

[0049]

[0050] Where M is the predicted value of the number of takeout orders in the calibrated forecast period, y is the ratio of the average number of takeout orders in the historical forecast period to the average number of total takeout orders per day, a is the average daily temperature on the forecast day, and a ′ is the historical daily average temperature, Δd t is the impact of daily average temperature on the number of takeout orders, b is the population density during the forecast period, and b ′ is the average population density during the historical forecast period, Δd d is the effect of population density on the number of takeout orders, M ′ is the forecast value of the number of takeout orders during a preset period. For example, Δd t Take 2, Δd d Take 3.

[0051] The embodiment of the present invention inputs the meteorological data and time data of the forecast period into the time-sharing order forecast model, outputs the forecast value of the number of takeout orders in the forecast period, and thus forecasts the number of takeout orders in each period. The forecast value of the number of takeout orders in the forecast period is calibrated using the calibration data to obtain the final forecast value of the number of takeout orders in the forecast period. The forecast value of the number of takeout orders in the forecast period is calibrated using the calibration data, so that the final forecast value of the number of takeout orders in the forecast period is more accurate, which is conducive to the platform to reasonably allocate the number of delivery riders.

[0052] The present invention also provides a method for training a time-sharing order prediction model for a takeaway platform. Figure 2 , as detailed below:

[0053] Step 201: Obtain historical takeaway order data from the takeaway platform, divide the 24 hours a day into n preset time periods, and summarize the number of takeaway orders in each preset time period based on the takeaway order data.

[0054] Extract historical order records from the food delivery platform’s database. These data usually include order time, order amount, user information, delivery address, payment method, etc.

[0055] Divide the 24 hours of a day into n time periods. These time periods can be of equal length, such as one time period per hour (0-1 o'clock, 1-2 o'clock, ..., 23-24 o'clock), or they can be customized, such as the morning peak (7-9 o'clock), the afternoon peak (11-13 o'clock), the evening peak (17-19 o'clock), etc.

[0056] Count the orders that occurred in each time period to get the number of orders in each period. This can be achieved by grouping and aggregating the order time in the order data.

[0057] The purpose of this process is to analyze the changes in order volume in different time periods, so as to help food delivery platforms or merchants understand which time periods of the day are peak order periods and which time periods are low order periods. Such analysis is very useful for optimizing operational strategies, adjusting human resource allocation, and improving marketing activities.

[0058] For example, if the order volume in a certain period is found to be significantly higher than in other periods, the food delivery platform may increase the number of delivery personnel during this period, or prompt the merchant to prepare more ingredients and packaging materials in advance to ensure that orders can be processed in time and improve customer satisfaction. At the same time, merchants can also adjust their promotional activities according to the fluctuation of order volume, such as providing discounts or coupons during periods with lower order volume to attract more customers to place orders.

[0059] Step 202: Obtain the meteorological data and time data for each preset time period, perform one-hot encoding on the meteorological data and time data, and standardize the meteorological data and time data using the min-max method so that the result value is mapped between [0-1].

[0060] The min-max method maps all numerical data (including meteorological data and numerical time data, such as hours) into the range of 0 to 1 in the process of standardizing meteorological data and time data, so that data of different dimensions can be compared and processed on the same scale.

[0061] In a possible implementation, before the meteorological data and the time data are one-hot encoded, the method further includes:

[0062] The missing data in the meteorological data are filled with meteorological data from nearby time periods, and outliers are removed.

[0063] Both missing values ​​and outliers may affect the results of data analysis and model prediction. By filling in missing values, the integrity of the data set can be maintained; by removing outliers, the quality of the data set can be improved, thereby improving the accuracy and reliability of subsequent analysis.

[0064] Step 203: Calculate the Pearson correlation coefficient between the weather data, time data and the number of takeout orders in each preset time period, retain the weather data and time data whose absolute value of the Pearson correlation coefficient is higher than the Pearson correlation coefficient threshold, and obtain optimized weather data and time data.

[0065] For example, the Pearson correlation coefficient threshold can be set to 0.9, and the time data and weather data with a Pearson correlation coefficient lower than the Pearson correlation coefficient threshold can be deleted, so as to obtain the time data and weather data with better correlation with the number of takeout orders, so as to make the data more representative and make the data of the time-sharing order quantity predicted by the present invention later more accurate.

[0066] For another example, the Pearson correlation coefficient threshold can be determined according to each Pearson correlation coefficient. The Pearson correlation coefficients are sorted according to size, and 10% of the time data and weather data with smaller Pearson correlation coefficients are deleted to obtain optimized weather data and time data.

[0067] In a possible implementation, the formula for calculating the Pearson correlation coefficient between the meteorological data, the time data, and the number of takeout orders in each preset time period is as follows:

[0068]

[0069] In the formula, ρ(x,y) is the Pearson correlation coefficient, x is the feature, including meteorological data and time data, and x i is the value of feature x at time period i, is the average value of feature x for all time periods, y is the number of takeaway orders for each preset time period, and y i is the number of takeaway orders in time period i, It is the average number of takeout orders in all time periods.

[0070] Through the Pearson correlation coefficient, we can understand the strength of the linear relationship between weather and time factors and the number of takeout orders. If the correlation coefficient is close to 1 or -1, it means there is a strong positive or negative correlation; if it is close to 0, it means there is no obvious linear relationship. This analysis helps food delivery platforms better understand the impact of different factors on order volume and make corresponding operational decisions.

[0071] Step 204: A time-sharing order prediction model is obtained based on the aggregated number of takeout orders in each preset time period and the optimized meteorological data and time data.

[0072] In a possible implementation, a time-sharing order prediction model is obtained based on the aggregated number of takeout orders in each preset time period and the optimized meteorological data and time data training, including:

[0073] The summarized number of takeout orders in each preset time period and the optimized meteorological data and time data are input into the extreme gradient boosting algorithm for training to obtain a time-sharing order prediction model.

[0074] In this embodiment, the extreme gradient boosting algorithm improves prediction accuracy by building multiple decision trees and optimizing them. This method performs well when processing data sets with complex feature relationships, so it is very suitable for tasks such as takeout order volume prediction, which may involve multiple influencing factors.

[0075] In a possible implementation, updating the weather data, time data, and the number of takeout orders in each preset time period at set intervals includes:

[0076] The newly added weather data, time data, and the number of takeout orders in each preset time period are summed with the original weather data, time data, and the number of takeout orders in each preset time period. The formula is as follows:

[0077] D updated =D geo +ΔD

[0078] Where D updated is the updated weather data, time data and the number of takeout orders in each preset time period, D geo is the original weather data, time data and the number of takeaway orders in each preset time period, and ΔD is the newly added weather data, time data and the number of takeaway orders in each preset time period.

[0079] In this embodiment, by continuously updating the meteorological data, time data and the number of takeout orders in each preset time period, the meteorological data, time data and the number of takeout orders in each preset time period are made more real-time, and further the predicted value of takeout orders in the predicted time period output by the time-sharing order prediction model is made more accurate and more real-time.

[0080] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0081] The following is an embodiment of the device of the present invention. For details not described in detail therein, reference may be made to the corresponding method embodiment described above.

[0082] Figure 3 The structure diagram of the device for predicting the number of time-sharing orders on a food delivery platform provided by an embodiment of the present invention is shown. For the sake of convenience, only the part related to the embodiment of the present invention is shown, which is described in detail as follows:

[0083] like Figure 3 As shown, the prediction device 3 for the number of time-sharing orders on the food delivery platform includes:

[0084] The time-sharing order prediction module 31 is used to input the meteorological data and time data of the prediction period into the time-sharing order prediction model, and output the predicted value of the takeaway order quantity in the prediction period;

[0085] The takeaway order quantity prediction value output module 32 is used to obtain calibration data, and calibrate the takeaway order quantity prediction value for the prediction period by using the calibration data to obtain the final takeaway order quantity prediction value for the prediction period.

[0086] The embodiment of the present invention inputs the meteorological data and time data of the forecast period into the time-sharing order forecast model, outputs the forecast value of the number of takeout orders in the forecast period, and thus forecasts the number of takeout orders in each period. The forecast value of the number of takeout orders in the forecast period is calibrated using the calibration data to obtain the final forecast value of the number of takeout orders in the forecast period. The forecast value of the number of takeout orders in the forecast period is calibrated using the calibration data, so that the final forecast value of the number of takeout orders in the forecast period is more accurate, which is conducive to the platform to reasonably allocate the number of delivery riders.

[0087] Figure 4 The structural diagram of the takeaway platform time-sharing order prediction model training device provided by the embodiment of the present invention is shown. For the convenience of explanation, only the part related to the embodiment of the present invention is shown, which is described in detail as follows:

[0088] like Figure 4 As shown, the takeaway platform time-sharing order prediction model training device 4 includes:

[0089] The takeaway order quantity aggregation module 41 for each preset time period obtains historical takeaway order data from the takeaway platform, divides 24 hours a day into n preset time periods, and aggregates the takeaway order quantity for each preset time period based on the takeaway order data;

[0090] The data standardization module 42 obtains the meteorological data and time data of each preset time period, performs one-hot encoding on the meteorological data and time data, and standardizes the meteorological data and time data by the min-max method so that the result value is mapped to between [0-1];

[0091] The data optimization module 43 calculates the Pearson correlation coefficient between the weather data, the time data and the number of takeout orders in each preset time period, retains the weather data and time data whose absolute value of the Pearson correlation coefficient is higher than the Pearson correlation coefficient threshold, and obtains optimized weather data and time data;

[0092] The time-sharing order prediction model training module 44 is used to train a time-sharing order prediction model based on the aggregated number of takeout orders in each preset time period and the optimized meteorological data and time data.

[0093] The embodiment of the present invention summarizes the number of takeout orders in each preset time period of the takeout platform to obtain the number of takeout orders in each preset time period in history, standardizes the meteorological data and time data of each preset time period, and facilitates the use of meteorological data and time data. By calculating the Pearson correlation coefficient between meteorological data, time data and the number of takeout orders in each preset time period, meteorological data and time data with low linear correlation are eliminated, so that the meteorological data, time data and the number of orders maintain a high linear correlation, and then the number of takeout orders in each preset time period and the optimized meteorological data and time data are input into the extreme gradient boosting algorithm. The extreme gradient boosting algorithm improves the prediction accuracy by constructing multiple decision trees and optimizing them. This method performs well in processing data sets with complex feature relationships, and is therefore very suitable for tasks such as takeout order volume prediction that may involve multiple influencing factors.

[0094] Figure 5 is a schematic diagram of a terminal provided by an embodiment of the present invention. Figure 5 As shown, the terminal 5 of this embodiment includes: a processor 50, a memory 51, and a computer program 52 stored in the memory 51 and executable on the processor 50. When the processor 50 executes the computer program 52, the steps in the above-mentioned embodiments of the method for predicting the number of time-sharing orders on a food delivery platform and the method for training a time-sharing order prediction model on a food delivery platform are implemented, for example Figure 1 , Figure 2 Alternatively, when the processor 50 executes the computer program 52, the functions of the modules in the above-mentioned device embodiments are realized, for example Figure 3 , Figure 4 The functions of each module are shown.

[0095] Exemplarily, the computer program 52 may be divided into one or more modules, which are stored in the memory 51 and executed by the processor 50 to implement the present invention. The one or more modules may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program 52 in the terminal 5. For example, the computer program 52 may be divided into Figure 3 , Figure 4 The modules shown.

[0096] The terminal 5 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The terminal 5 may include, but is not limited to, a processor 50 and a memory 51. Those skilled in the art will appreciate that Figure 5 It is only an example of terminal 5 and does not constitute a limitation on terminal 5. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal may also include input and output devices, network access devices, buses, etc.

[0097] The processor 50 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0098] The memory 51 may be an internal storage unit of the terminal 5, such as a hard disk or memory of the terminal 5. The memory 51 may also be an external storage device of the terminal 5, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal 5. Further, the memory 51 may also include both an internal storage unit of the terminal 5 and an external storage device. The memory 51 is used to store the computer program and other programs and data required by the terminal. The memory 51 may also be used to temporarily store data that has been output or is to be output.

[0099] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0100] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0101] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0102] In the embodiments provided by the present invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0103] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0104] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0105] If the integrated module is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, it can implement the steps of the above-mentioned various food delivery platform time-sharing order quantity prediction methods and food delivery platform time-sharing order prediction model training method embodiments. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practices in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practices, computer-readable media does not include electrical carrier signals and telecommunication signals.

[0106] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for predicting the number of time-sharing orders on a food delivery platform, characterized in that: include: The meteorological data and time data of the forecast period are input into the time-sharing order forecasting model, and the forecast value of the number of takeout orders in the forecast period is output; The calibration data is obtained, and the prediction value of the number of takeout orders for the prediction period is calibrated by the calibration data to obtain the final prediction value of the number of takeout orders for the prediction period.

2. The method for predicting the number of time-sharing orders on a food delivery platform according to claim 1, characterized in that: The calibration data includes: The ratio of the average number of takeout orders in each historical period to the average number of total daily takeout orders, population density, and temperature.

3. The method for predicting the number of time-sharing orders on a food delivery platform according to claim 2, characterized in that: The step of calibrating the predicted value of the number of takeout orders for the prediction period by using the calibration data includes: Based on the impact of changes in population density and temperature on the number of takeout orders, the ratio of the average number of takeout orders in each historical period to the average number of daily takeout orders is used to calibrate the forecast value of the number of takeout orders in the forecast period. The formula is: M=y[|a-a ′ |Δd a +|b-b ′ |Δd b ]M ′ Where M is the predicted value of the number of takeaway orders in the calibrated forecast period, y is the ratio of the average number of takeaway orders in the historical forecast period to the average number of total takeaway orders per day, a is the average daily temperature on the forecast day, and a ′ is the historical daily average temperature, Δd a is the weight coefficient of the daily average temperature on the takeaway order volume, b is the population density during the forecast period, and b ′ is the average population density during the historical forecast period, Δd b is the weight coefficient of population density on the number of takeout orders, M ′ It is the predicted value of the number of takeout orders for the preset time period.

4. A method for training a time-sharing order prediction model for a food delivery platform, characterized in that: include: Obtain historical takeaway order data from the takeaway platform, divide the 24 hours a day into n preset time periods, and summarize the number of takeaway orders in each preset time period based on the takeaway order data; Obtain the meteorological data and time data for each preset time period, perform one-hot encoding on the meteorological data and time data, and standardize the meteorological data and time data using the min-max method so that the result value is mapped between [0-1]; Calculate the Pearson correlation coefficient between the weather data, time data and the number of takeout orders in each preset time period, retain the weather data and time data whose absolute value of the Pearson correlation coefficient is higher than the Pearson correlation coefficient threshold, and obtain optimized weather data and time data; The time-sharing order prediction model is obtained based on the summarized number of takeout orders in each preset time period and the optimized meteorological data and time data training.

5. The method for training a time-sharing order prediction model for a food delivery platform according to claim 4, characterized in that: Before the meteorological data and the time data are uniquely encoded, the method further includes: The missing data in the meteorological data are filled with meteorological data from nearby time periods, and outliers are removed.

6. The method for training a time-sharing order prediction model for a food delivery platform according to claim 4, characterized in that: The formula for calculating the Pearson correlation coefficient of meteorological data, time data and the number of takeout orders in each preset time period is as follows: In the formula, ρ(x,y) is the Pearson correlation coefficient, x is the feature, including meteorological data and time data, and x i is the value of feature x at time period i, is the average value of feature x for all time periods, y is the number of takeaway orders for each preset time period, and y i is the number of takeaway orders in time period i, It is the average number of takeout orders in all time periods.

7. The method for training a time-sharing order prediction model for a food delivery platform according to claim 4, characterized in that: Update weather data, time data and the number of takeout orders for each preset time period at set intervals, including: The newly added weather data, time data, and the number of takeout orders in each preset time period are summed with the original weather data, time data, and the number of takeout orders in each preset time period. The formula is as follows: D updated =D geo +ΔD Where D updated is the updated weather data, time data and the number of takeout orders in each preset time period, D geo is the original weather data, time data and the number of takeaway orders in each preset time period, and ΔD is the newly added weather data, time data and the number of takeaway orders in each preset time period.

8. The method for training a time-sharing order prediction model for a food delivery platform according to claim 4, characterized in that: The time-sharing order prediction model is obtained by training the aggregated takeout order quantity in each preset time period and the optimized meteorological data and time data, including: The summarized number of takeout orders in each preset time period and the optimized meteorological data and time data are input into the extreme gradient boosting algorithm for training to obtain a time-sharing order prediction model.

9. A terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method as claimed in any one of claims 1 to 8 are implemented.

Images