Demand prediction method and device based on service feedback, and storage medium

Abstract

The invention discloses a demand prediction method and device based on service feedback, electronic equipment and a storage medium, and relates to the technical field of demand prediction of supply chain management. The method comprises the following steps: acquiring basic data, and cleaning and preprocessing the basic data; based on the basic data, fusing service feedback information, and establishing a service experience database; calling data in the business experience database, completing variable analysis and aggregation through standardized and derived variables, and obtaining a first prediction model through variable screening, parameter traversal and training; adjusting data of the business experience database based on business update data, and optimizing parameters of the first prediction model to obtain a second prediction model; and generating demand data in a predetermined time period by using the second prediction model. The invention also discloses a demand prediction device based on business feedback, electronic equipment and a storage medium, so as to effectively integrate experience of business personnel, iteratively optimize the prediction model and establish a forward feedback mechanism.

Description

Technical Field

[0001] The present application relates to the technical field of demand forecasting for supply chain management, and in particular to a demand forecasting method, device and storage medium based on business feedback. Background Art

[0002] In the field of supply chain management, demand forecasting plays a vital role, and its core task is to estimate the demand trend of the future market. This process is crucial for enterprises because it is directly related to the formulation of production plans, the control of inventory levels, and the optimization of sales strategies, all of which are aimed at better meeting customer needs and maximizing profits. The accuracy of demand forecasting results has a significant impact on many aspects of enterprise operations, including production efficiency, inventory costs, and customer satisfaction.

[0003] In traditional demand forecasting practices, commonly used methods include time series analysis, statistical models, and machine learning algorithms. Time series analysis predicts future trends by studying the sequence patterns of historical data, while statistical models rely on mathematical and statistical principles to estimate demand. Machine learning algorithms predict future demand changes by learning patterns from historical data. With the continuous advancement of big data and artificial intelligence technologies, the application of deep learning technology in the field of demand forecasting is also increasing. Deep learning models, especially those complex models that can process large amounts of historical data, have begun to show their great potential in capturing complex patterns in data and improving forecast accuracy.

[0004] However, demand data in the supply chain is often the result of the interaction between the actual market situation and multiple factors such as internal production planning and inventory management. These data not only reflect the dynamic changes in the market, but are also affected by the internal decisions of the enterprise and changes in the external market environment. In addition, enterprises often face challenges in obtaining real data on the market side, which may be difficult to obtain or delayed.

[0005] Therefore, both traditional demand forecasting algorithms and deep learning-based technologies may encounter challenges in forecasting accuracy in practical applications. Business personnel may question the forecasting results of the model, especially when demand data is affected by multiple complex factors. Improving the accuracy and credibility of the forecasting model and forming a positive cycle between artificial intelligence algorithms and business personnel has always been a major pain point in the integration of technology and business. Summary of the invention

[0006] The embodiments of the present application provide a demand forecasting method, device, electronic device and storage medium based on business feedback, which are used to solve the problems in existing demand forecasting methods, such as failure to effectively integrate the experience of business personnel, insufficient iterative optimization of forecasting models, and lack of positive feedback mechanism.

[0007] On the one hand, an embodiment of the present application provides a demand forecasting method based on business feedback, including:

[0008] Obtain basic data and perform cleaning and preprocessing;

[0009] Based on the basic data, business feedback information is integrated to establish a business experience database;

[0010] Calling the data in the business experience database, completing variable analysis and aggregation through standardized and derived variables, and training to obtain a first prediction model through variable screening and parameter traversal;

[0011] Based on the business update data, adjusting the data of the business experience database, optimizing the parameters of the first prediction model, and obtaining a second prediction model;

[0012] The second prediction model is used to generate demand data within a predetermined time period.

[0013] In a possible embodiment, basic data is obtained and cleaned and preprocessed, including: traversing and searching the lower links involved in the predetermined website according to preset parameters, and crawling the basic data of the predetermined website, wherein the preset parameters include the address of the predetermined website, the traversal depth value and the crawling cycle, and the basic data includes historical sales data, market trends and promotional activities information; cleaning the basic data, the cleaning operation includes storing incomplete data, erroneous data or redundant data in a redundant data set for backup; preprocessing the basic data and storing it in a central database, the preprocessing operation includes describing the type, attribute and function of the data.

[0014] In a possible embodiment, based on the basic data, business feedback information is integrated to establish a business experience database, including: backing up the basic data to a first data table; optimizing query statements in the storage process of the business experience database, including sub-query statements therein; sorting the real tables and virtual tables involved in the query statements according to the size and query rate, and arranging them in order from small to large in order of right first, then left, and bottom first, then top; establishing database indexes for fields related to historical sales data, market trends and promotional information through database scripts, so as to speed up the adjustment of data values ​​of related fields when business information is updated; importing data from the first data table into the second data table, so that the first data table can be deleted while ensuring that the business can be executed.

[0015] In a possible embodiment, the data in the business experience database is called, and variable analysis and aggregation are completed through standardization and derived variables, and a model is trained through variable screening and parameter traversal, including: obtaining the standard field name, data format, and standard field value range of the data in the business experience database; completing variable analysis and aggregation of the derived variables, and inputting normalization processing to generate a unified format for subsequent input into the first prediction model training; constructing a random forest model as the first prediction model to predict market demand, including the following specific steps: first, using the historical sales data of the basic data to form a training sample, wherein the number of samples of the training sample is recorded as N, and N is a positive integer; second, k training sets are randomly taken out from the training sample with replacement, and k≤N, and a corresponding regression decision tree is created for each training set to form a random forest model containing k decision trees; third, a part of the training sample is randomly selected again as the first test set, and the first test set data is input into the random forest model for testing, and the parameters of the random forest model are traversed for optimization.

[0016] In a possible embodiment, based on business update data, the data in the business experience database is adjusted to optimize the parameters of the first prediction model to obtain a second prediction model, including: obtaining the business update data, wherein the business update data includes at least one of the following: historical sales data, market trends or promotional activity information; merging the business update data with the same type of data in the business experience database as a second test set; and using the second test set to optimize the parameters of the first prediction model to obtain the second prediction model.

[0017] In a possible embodiment, the parameters of the first prediction model are optimized using the second test set to obtain the second prediction model, including: using the second test set to preliminarily evaluate the score of the second prediction model, and adjusting the parameters based on the model score, adjusting the maximum number of iterations of the weak learner and the decision tree depth model fitting parameters; using the evaluation indicators mean absolute error, root mean square error, and mean absolute percentage error to evaluate the model prediction effect.

[0018] In a possible embodiment, the method further includes: using a graphical user interface to guide input of the service feedback information and / or the service update data.

[0019] On the one hand, an embodiment of the present application provides a demand forecasting device based on business feedback, including:

[0020] Data acquisition module, used to obtain basic data and perform cleaning and preprocessing;

[0021] A database module is used to establish a business experience database based on the basic data and integrating business feedback information;

[0022] A model training module is used to call the data in the business experience database, complete variable analysis and aggregation through standardized and derived variables, and train a first prediction model through variable screening and parameter traversal;

[0023] A model optimization module, used to adjust the data of the business experience database based on the business update data, optimize the parameters of the first prediction model, and obtain a second prediction model;

[0024] A result generation module is used to generate demand data within a predetermined time period using the second prediction model.

[0025] On the one hand, an embodiment of the present application provides an electronic device, which includes a processor and a memory, wherein the memory stores program code, and when the program code is executed by the processor, the processor executes any of the above-mentioned demand forecasting methods based on business feedback.

[0026] On the one hand, the present application provides a computer-readable storage medium, which includes a program code. When the storage medium is run on an electronic device, the program code is used to enable the electronic device to execute any of the above-mentioned demand forecasting methods based on business feedback.

[0027] The beneficial effects brought by the technical solution provided by the embodiment of the present application include at least one of the following:

[0028] 1. Improve forecast accuracy: By combining the experience of business personnel with the algorithm model, this application can improve the accuracy of demand forecasting. Business personnel can integrate business feedback information, establish a business experience database, and combine their deep understanding of market dynamics with the algorithm model's ability to analyze historical data, making the forecast results closer to actual market demand.

[0029] 2. Enhance model adaptability: The solution of this application not only establishes a business experience database by continuously feeding back adjustments made by business personnel, but also continuously uses business update data based on actual conditions, so that the algorithm model can adapt to market changes, thereby improving the flexibility and adaptability of the model.

[0030] 3. Increase data assets: business experience database and business update data, which can accumulate the human functional experience of sales staff.

[0031] 4. Establish a communication bridge between front-end sales staff and back-end data: Based on the data in the business experience database and integrated with the business update data, accurate prediction results can help enterprises better plan resources, from enterprise planning, marketing, inventory management, production planning, to pull together data cognition and build a data bridge, thereby improving operational efficiency.

[0032] 5. Improve decision support: This application uses the demand data within a predetermined time period generated by the model. The forecast results provided can serve as an important basis for corporate decision-making and help management make more scientific strategic planning and market response.

[0033] 6. Achieve continuous learning and iteration: The solution of this application supports continuous learning of the model, using data from the business experience database and optimizing the model; thus, through continuous optimization and iteration, the prediction model can improve itself over time.

[0034] 7. Enhance user experience: Optimizing models and updating business data can obtain more accurate predictions, so that companies can better meet customer needs and improve customer satisfaction and loyalty.

[0035] Other features and advantages of the present application will be described in the following description, and partly become apparent from the description, or understood by practicing the present application. The purpose and other advantages of the present application can be realized and obtained by the structures specifically pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0037] Figure 1 A schematic diagram of an application scenario of the demand forecasting method based on business feedback provided in an embodiment of the present application;

[0038] Figure 2 A flow chart of a demand forecasting method based on business feedback provided in an embodiment of the present application;

[0039] Figure 3 A schematic diagram of the working process of the demand forecasting method based on business feedback provided in an embodiment of the present application;

[0040] Figure 4 A module diagram of a demand forecasting device based on business feedback provided in an embodiment of the present application;

[0041] Figure 5 The present invention is a schematic diagram of a hardware structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical scheme and advantages of the present application clearer, the technical scheme in the embodiment of the present application will be clearly and completely described below in conjunction with the drawings in the embodiment of the present application. Obviously, the described embodiment is only a part of the embodiment of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application. In the absence of conflict, the embodiments in the present application and the features in the embodiments can be arbitrarily combined with each other. In addition, although the logical order is shown in the flow chart, in some cases, the steps shown or described can be performed in an order different from that here.

[0043] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0044] The following is a brief introduction to the design concept of the embodiment of the present application:

[0045] In today's enterprise demand forecasting practice, despite combining advanced algorithm models and the rich experience of business experts to improve the accuracy of forecasts, there are still several challenges: First, the utilization of historical experience is insufficient: the adjustments and suggestions of business experts on forecast results usually only affect the current forecast output, but fail to be systematically recorded and integrated into the subsequent forecasting process. This fragmented processing method leads to the failure to effectively reuse valuable historical experience and related data, thereby hindering the ability of the forecasting process to learn and evolve from past corrections. Second, the lack of quantification of manual experience: In the industrial world, due to the lack of comprehensive grasp of terminal and upstream data, information transmission often relies on verbal communication of business personnel, and the information obtained from other companies is often highly aggregated and condensed data. This leads to challenges in the standardization and structuring of business knowledge, and it is difficult to convert these intangible assets into quantifiable forecasting inputs. Third, the lack of iterative optimization of models: the historical experience and market insights of business personnel often fail to be fed back to the source of model development and training. This disconnection hinders the model from being optimized and iterated again based on feedback from actual applications, limiting the continuous improvement of model performance. Fourth, the lack of positive feedback mechanism: There is a lack of effective positive feedback and interaction mechanism between the prediction results of the algorithm model and the manual experience of business personnel. Ideally, the adjustments made by business personnel should guide model learning, and the prediction results of the model should also assist business personnel in gaining a deeper understanding of market dynamics. However, existing processes often fail to establish such a mutually optimized and iterative relationship, resulting in limited prediction accuracy and the ability to respond to market changes.

[0046] In view of this, the embodiments of the present application provide a demand forecasting method, device, electronic device and storage medium based on business feedback. Among them, the demand forecasting method based on business feedback includes: obtaining basic data, cleaning and preprocessing; based on the basic data, integrating business feedback information, and establishing a business experience database; calling the data in the business experience database, completing variable analysis and aggregation through standardized and derived variables, and training to obtain a first forecasting model through variable screening and parameter traversal; based on business update data, adjusting the data in the business experience database, optimizing the parameters of the first forecasting model, and obtaining a second forecasting model; using the second forecasting model to generate demand data within a predetermined time period.

[0047] The preferred embodiments of the present application are described below in conjunction with the drawings in the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application. In addition, the embodiments and features in the embodiments of the present application may be combined with each other if there is no conflict.

[0048] like Figure 1, which is a schematic diagram of an application scenario of a demand forecasting method based on business feedback provided by an embodiment of the present application. In the schematic diagram of the application scenario, a terminal device 101 and a server 102 are included. The terminal device 101 and the server 102 communicate with each other through a communication network.

[0049] The terminal device 101 is an electronic device used by the target object, and the electronic device may be a personal computer, a mobile phone, a tablet computer, a notebook, an e-book reader, a vehicle-mounted terminal, etc. In addition, a client related to the demand forecasting method based on business feedback may be installed on the terminal device 101, and the client may be software (for example, an APP, a browser, etc.), or a web page, a small program, etc. By implementing the above-mentioned demand forecasting method based on business feedback on the terminal device 101, a demand forecasting operation related to business feedback is performed.

[0050] The server 102 may be an independent physical server or an edge device 102 in the field of cloud computing. It may also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud storage, cloud functions, network services, cloud communications, middleware services, domain name services, security services, content distribution networks (English name: Content Delivery Network, abbreviated as CDN), as well as big data and artificial intelligence platforms.

[0051] There is no restriction on the number of the terminal devices 101 and / or servers 102 .

[0052] It should be noted that the demand prediction method based on business feedback in the embodiment of the present application can be executed by the terminal device 101 or the server 102 alone, or can be executed jointly by the terminal device 101 and the server 102. For example, when executed jointly by the terminal device 101 and the server 102, the terminal device 101 can provide basic data to the server 102, and perform cleaning and preprocessing; the server 102 can integrate business feedback information based on the basic data to establish a business experience database. The server 102 calls the data in the business experience database, completes variable analysis and aggregation through standardization and derived variables, and obtains a first prediction model through variable screening and parameter traversal through training; based on the business update data, adjusts the data in the business experience database, optimizes the parameters of the first prediction model, and obtains a second prediction model; uses the second prediction model to generate demand data within a predetermined time period. The terminal device 101 obtains the demand data within a predetermined time period generated by the second prediction model from the server 102.

[0053] The following describes the warehouse site selection method provided by the exemplary embodiment of the present application in combination with the above-mentioned application scenarios and with reference to the accompanying drawings. It should be noted that the above-mentioned application scenarios are only shown to facilitate understanding of the spirit and principles of the present application, and the implementation methods of the present application are not subject to any limitations in this regard.

[0054] like Figure 2 , which is a schematic diagram of a demand forecasting method 200 based on business feedback provided in an embodiment of the present application. In the schematic diagram of the demand forecasting method based on business feedback, the following steps are shown:

[0055] Step S201, obtaining basic data, that is, obtaining basic data, and performing cleaning and preprocessing.

[0056] In a possible embodiment, basic data is obtained, and cleaned and preprocessed, including: traversing the lower-level links involved in the predetermined website according to preset parameters to perform traversal and search, and crawling the basic data of the predetermined website, wherein the preset parameters include the address of the predetermined website, the traversal depth value and the crawling cycle, and the basic data includes historical sales data, market trends and promotional activities information; cleaning the basic data, the cleaning operation includes storing incomplete data, erroneous data or redundant data in a redundant data set for standby, for example: data on non-sales days such as store construction and decoration days in historical sales data can be eliminated, that is, redundant data can be eliminated; preprocessing the basic data, storing it in a central database, the preprocessing operation includes describing the type, attributes and functions of the data. For example, it is used to crawl the URL (Uniform Resource Location, web page address) of the website according to preset parameters, parse DOM data that supports Jquery syntax, parse json data and / or parse script data, parse and extract information (including data information, pictures and / or flash and other binary files). In particular, the preset parameters may include: an address of a predetermined website, a traversal depth value, and a crawling cycle.

[0057] Use preset parameters to traverse the predetermined websites. You can set the preset parameters according to the system's initialization environment, server performance, network bandwidth, and the number of crawling processes. Set the number of processes and websites to be crawled concurrently, and distribute the crawling processes evenly to each website. This can both reasonably crawl the websites and avoid the pressure of intensive visits by crawlers to the crawled websites, while maintaining efficiency.

[0058] Step S202, establish a business experience database, that is, based on the basic data, integrate business feedback information to establish a business experience database. Integration of business feedback information can improve prediction accuracy. By combining the experience of business personnel with the algorithm model, this embodiment can improve the accuracy of demand forecasting. The business personnel's deep understanding of market dynamics is combined with the algorithm model's ability to analyze historical data, making the prediction results closer to actual market demand. The solution of the present application increases data assets and precipitates the human functional experience of business personnel by continuously feeding back adjustments to business personnel; enabling the algorithm model to adapt to market changes, thereby improving the flexibility and adaptability of the model.

[0059] In a possible embodiment, based on the basic data, business feedback information is integrated to establish a business experience database, including: backing up the basic data to a first data table; optimizing query statements in the storage process of the business experience database, including sub-query statements therein; sorting the real tables and virtual tables involved in the query statements according to the size and query rate, and arranging them in order from small to large in order of right first, then left, and bottom first, then top; establishing database indexes for fields related to historical sales data, market trends and promotional information through database scripts, so as to speed up the adjustment of data values ​​of related fields when business information is updated; importing data from the first data table into the second data table, so that the first data table can be deleted while ensuring that the business can be executed. For example, you can use the ORACLE database system and the WHERE clause of the query statement to extract all the conditions one by one and perform query tests in the actual environment; sort the query rates and optimize the WHERE clauses in order of right to left and bottom to top from small to large; and record the query filter conditions that take the longest time and the query filter conditions that are used most frequently; from the above steps, it can be concluded that the most frequently occurring and time-consuming search conditions include fields of historical sales data, market trends, and promotional activity information: the sales volume of historical sales data SALE_DATA, the price of historical sales data SALE_PRICE, the market trend MARKET_PROPERTY, and the promotional activity information MARKET_ACCU. Database indexes are established for the fields SALE_DATA, SALE_PRICE, MARKET_PROPERTY, and MARKET_ACCU through database scripts to improve the execution efficiency of the business experience database.

[0060] Step S203, training to obtain a first prediction model, that is, calling the data in the business experience database, completing variable analysis and aggregation through standardized and derived variables, and training to obtain a first prediction model through variable screening and parameter traversal.

[0061] In a possible embodiment, the data in the business experience database is called, and variable analysis and aggregation are completed through standardization and derived variables, and a model is trained through variable screening and parameter traversal, including: obtaining the standard field name, data format, and standard field value range of the data in the business experience database; completing variable analysis and aggregation of the derived variables, and inputting normalization processing to generate a unified format for subsequent input into the first prediction model training; constructing a random forest model as the first prediction model to predict market demand, including the following specific steps: first, using the historical sales data of the basic data to form a training sample, wherein the number of samples of the training sample is recorded as N, and N is a positive integer; second, k training sets are randomly taken out from the training sample with replacement, and k≤N, and a corresponding regression decision tree is created for each training set to form a random forest model containing k decision trees; third, a part of the training sample is randomly selected again as the first test set, and the first test set data is input into the random forest model for testing, and the parameters of the random forest model are traversed for optimization.

[0062] For example, we can assume that the historical data of the first three months of the same period within two years are taken out to form the training sample D, where the last data is the input data used for prediction. Assume that the number of variables selected by each regression decision tree is represented by a numerical value. Randomly take out k (the value can be 200) training sets from the training sample D with replacement, and each training set contains N data. Create a regression decision tree for each training set to form a random forest model with k decision trees. The last recent data, that is, the M-dimensional sales feature data of this month, is input into the random forest model to obtain the output results of all regression decision trees to verify the model function.

[0063] Table 1 shows the characteristic value information of sales data for the design and verification of random forest model:

[0064] F0 F1 F2 F3 …… F15 1 7 8555.3 8055.7 8092.1 …… 1011.8 2 5.8 8102.6 8024.9 8053.9 …… 1075.2 3 42.22 8755.1 8183.8 8154.7 …… 1101.4 …… …… …… …… …… …… ……

[0065] Table 1: Sales data characteristic value information

[0066] According to Table 1, we can specify a number 6, which is much smaller than the number of features 16, as the number of input features. 100 regression decision trees are built. The training set of each decision tree is a decision tree built by randomly taking 16 samples from 16 data with replacement. The sales feature data of the first test set data is input into the decision tree to get the predicted value of the demand data.

[0067] Step S204, optimize the first prediction model to obtain a second prediction model, that is, based on the business update data, adjust the data of the business experience database, optimize the parameters of the first prediction model, and obtain the second prediction model. The embodiments disclosed in this application support continuous learning of the model, and through continuous optimization and iteration, the prediction model can improve itself over time.

[0068] In a possible embodiment, based on business update data, the data in the business experience database is adjusted to optimize the parameters of the first prediction model to obtain a second prediction model, including: obtaining the business update data, wherein the business update data includes at least one of the following: historical sales data, market trends or promotional activity information; merging the business update data with the same type of data in the business experience database as a second test set; and using the second test set to optimize the parameters of the first prediction model to obtain the second prediction model.

[0069] Step S205, generating demand data within a predetermined time period, that is, using the second forecasting model to generate demand data within a predetermined time period. The forecasting results provided by this embodiment can serve as an important basis for enterprise decision-making, helping management to make more scientific strategic planning and market response. Accurate forecasting results can help enterprises better plan resources, from enterprise planning, marketing, inventory management, production planning, to pull through data cognition, and see a data bridge, thereby improving operational efficiency.

[0070] In a possible embodiment, the parameters of the first prediction model are optimized using the second test set to obtain the second prediction model, including: using the second test set to preliminarily evaluate the score of the second prediction model, and adjusting the parameters based on the model score, adjusting the maximum number of iterations of the weak learner and the decision tree depth model fitting parameters; using the evaluation indicators mean absolute error, root mean square error, and mean absolute percentage error to evaluate the model prediction effect.

[0071] In a possible embodiment, the demand forecasting method based on business feedback may further include: using a graphical user interface to guide the input of the business feedback information and / or the business update data. By designing a user interface prototype, the graphical user interface is ensured to be intuitive and easy to use. By developing a front-end interface, business personnel can easily input adjustment suggestions and view forecast results. In addition, business personnel can use the graphical user interface to interact with the business experience database, such as: data query request.

[0072] Figure 3A schematic diagram of the working process of the demand forecasting method based on business feedback is shown. Basic data 301 may include historical sales data, market trends and promotional information; establishing a business experience database 302, the URL (Uniform Resource Location, web address) of the website may be captured by preset parameters, that is, using Internet data capture 303, parsing DOM data that supports Jquery-like syntax, parsing json data and / or parsing data that supports script scripts, parsing and extracting information (including data information, pictures and / or flash and other binary files); establishing a model 304, a random forest model may be constructed to predict market demand; an interactive user interface 305 may use a graphical user interface to guide the input of the business feedback information and / or the business update data; so that business personnel can use the graphical user interface to interact with the business experience database; feedback business data 306 may be based on the graphical user interface or directly change the data in the business experience database; adaptive learning of the model 307, merging the business update data with the same type of data in the business experience database to achieve multi-dimensional data fusion as a second test set; then, based on the second test set, the adaptive learning of the model is used to adjust the corresponding parameters to optimize the random forest model.

[0073] refer to Figure 4 , which is a module diagram of a demand forecasting device based on business feedback provided in an embodiment of the present application, with reference to Figure 4 The device includes: a data acquisition module 402, a database module 404, a model training module 406, a model optimization module 408 and a result generation module 410. Among them:

[0074] The data acquisition module 402 is used to acquire basic data and perform cleaning and preprocessing.

[0075] The database module 404 is used to establish a business experience database based on the basic data and integrating the business feedback information.

[0076] The model training module 406 is used to call the data in the business experience database, complete variable analysis and aggregation through standardization and derived variables, and obtain the first prediction model through variable screening and parameter traversal.

[0077] The model optimization module 408 is used to adjust the data of the business experience database based on the business update data, optimize the parameters of the first prediction model, and obtain the second prediction model.

[0078] The result generating module 410 is used to generate demand data within a predetermined time period using the second prediction model.

[0079] Based on the same inventive concept, an electronic device is also provided in the embodiment of the present application, and the electronic device can implement the aforementioned demand prediction method based on business feedback, referring to Figure 5 , the electronic device comprises:

[0080] At least one processor 501, and a memory 502 connected to the at least one processor 501. The specific connection medium between the processor 501 and the memory 502 is not limited in the embodiment of the present application. Figure 5 In the example, the processor 501 and the memory 502 are connected via a bus 500. The bus 500 is Figure 5 The connections between other components are shown in bold lines, and are not intended to be limiting. The bus 500 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 Only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. Alternatively, the processor 501 can also be called a controller, and there is no limitation on the name.

[0081] In the embodiment of the present application, the memory 502 stores instructions that can be executed by at least one processor 501. The at least one processor 501 can execute the demand prediction method based on business feedback discussed above by executing the instructions stored in the memory 502. The processor 501 can implement Figure 4 The functions of each module in the device shown.

[0082] Among them, the processor 501 is the control center of the device, and can use various interfaces and lines to connect the various parts of the entire control device. By running or executing instructions stored in the memory 502 and calling data stored in the memory 502, the various functions of the device and processing data, the device can be monitored as a whole.

[0083] In one possible design, the processor 501 may include one or more processing units, and the processor 501 may integrate an application processor and a modem processor, wherein the application processor mainly processes an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the modem processor may not be integrated into the processor 501. In some embodiments, the processor 501 and the memory 502 may be implemented on the same chip, and in some embodiments, they may also be implemented separately on separate chips.

[0084] Processor 501 can be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, and can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. A general-purpose processor can be a microprocessor or any conventional processor, etc. The steps of the demand forecasting method based on business feedback disclosed in the embodiments of the present application can be directly embodied as a hardware processor execution, or a combination of hardware and software modules in the processor.

[0085] The memory 502 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules. The memory 502 may include at least one type of storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory, a random access memory (Random Access Memory, RAM), a static random access memory (Static Random Access Memory, SRAM), a programmable read-only memory (Programmable Read Only Memory, PROM), a read-only memory (Read Only Memory, ROM), an electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), a magnetic memory, a disk, an optical disk, etc. The memory 502 is any other medium that can be used to carry or store a desired program code in the form of an instruction or data structure and can be accessed by a computer, but is not limited thereto. The memory 502 in the embodiment of the present application can also be a circuit or any other device that can realize a storage function, for storing program instructions and / or data.

[0086] By programming the processor 501, the code corresponding to the demand prediction method based on business feedback introduced in the above embodiment can be fixed into the chip, so that the chip can execute the code when running. Figure 2 The steps of the demand prediction method based on business feedback in the illustrated embodiment. How to design and program the processor 501 is a technology well known to those skilled in the art and will not be described in detail here.

[0087] Based on the same inventive concept, an embodiment of the present application further provides a storage medium storing computer instructions. When the computer instructions are executed on a computer, the computer executes the demand forecasting method based on business feedback discussed above.

[0088] In some possible implementations, various aspects of the demand forecasting method based on business feedback provided in the present application can also be implemented in the form of a program product, which includes a program code. When the program product is run on an apparatus, the program code is used to enable the control device to execute the steps of the demand forecasting method based on business feedback according to various exemplary implementations of the present application described above in this specification.

[0089] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0090] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0091] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0092] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0093] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A demand forecasting method based on business feedback, characterized in that: include: Obtain basic data and perform cleaning and preprocessing; Based on the basic data, business feedback information is integrated to establish a business experience database; Calling the data in the business experience database, completing variable analysis and aggregation through standardized and derived variables, and training to obtain a first prediction model through variable screening and parameter traversal; Based on the business update data, adjusting the data in the business experience database, optimizing the parameters of the first prediction model, and obtaining a second prediction model; The second prediction model is used to generate demand data within a predetermined time period.

2. The demand forecasting method based on business feedback as claimed in claim 1, characterized in that: Obtain basic data and perform cleaning and preprocessing, including: Traversing the lower links involved in the predetermined website according to preset parameters to perform traversal search, and crawling the basic data of the predetermined website, wherein the preset parameters include the address of the predetermined website, the traversal depth value and the crawling cycle, and the basic data includes historical sales data, market trends and promotional activity information; Performing a cleaning operation on the basic data, wherein the cleaning operation includes storing incomplete data, erroneous data or redundant data in a redundant data set for standby use; The basic data is preprocessed and stored in a central database, wherein the preprocessing operation includes describing the type, attribute and function of the data.

3. The demand forecasting method based on business feedback as claimed in claim 1, characterized in that: Based on the basic data, business feedback information is integrated to establish a business experience database, including: Backing up the basic data to a first data table; Optimize the query statements in the storage process of the business experience database, including the subquery statements therein; sort the real tables and virtual tables involved in the query statements according to the size and query rate, and arrange them in order from small to large in the order of right first, then left, and bottom first, then top; Use database scripts to create database indexes for fields related to historical sales data, market trends, and promotional information, to speed up the adjustment of data values ​​in related fields when business information is updated; The data in the first data table is imported into the second data table, so that the first data table is deleted while ensuring that the business can be executed.

4. The demand forecasting method based on business feedback as claimed in claim 1, characterized in that: The data in the business experience database is called, and variable analysis and aggregation are completed through standardized and derived variables. The model is trained through variable screening and parameter traversal, including: Obtaining the standard field name, data format, and standard field value range of the data in the business experience database; Performing variable analysis and aggregation on the derived variables, and inputting normalization processing to generate a unified format for subsequent input into the first prediction model training; Constructing a random forest model as the first prediction model to predict market demand includes the following specific steps: first, using the historical sales data of the basic data to form a training sample, wherein the number of samples of the training sample is recorded as N, and N is a positive integer; second, randomly taking out k training sets from the training sample with replacement, and k≤N, creating a corresponding regression decision tree for each training set, and forming a random forest model containing k decision trees; third, randomly selecting a part of the training sample again as the first test set, inputting the first test set data into the random forest model for testing, and traversing the parameters of the random forest model for optimization.

5. The demand forecasting method based on business feedback as claimed in claim 1, characterized in that: Based on the business update data, adjusting the data of the business experience database, optimizing the parameters of the first prediction model, and obtaining the second prediction model, including: Acquiring the business update data, wherein the business update data includes at least one of the following: historical sales data, market trends or promotional activity information; Merging the business update data with the same type of data in the business experience database as a second test set; The parameters of the first prediction model are optimized using the second test set to obtain the second prediction model.

6. The demand forecasting method based on business feedback as claimed in claim 5, characterized in that: Optimizing the parameters of the first prediction model using the second test set to obtain the second prediction model includes: Preliminarily evaluating the score of the second prediction model using the second test set, and adjusting parameters based on the model score, adjusting the maximum number of iterations of the weak learner and the decision tree depth model fitting parameters; The evaluation indicators mean absolute error, root mean square error, and mean absolute percentage error were used to evaluate the model prediction effect.

7. The demand forecasting method based on business feedback according to any one of claims 1 to 6, characterized in that: Also includes: The graphical user interface is used to guide the input of the service feedback information and / or the service update data.

8. A demand forecasting device based on business feedback, characterized in that: include: Data acquisition module, used to obtain basic data and perform cleaning and preprocessing; A database module is used to establish a business experience database based on the basic data and integrating business feedback information; A model training module is used to call the data in the business experience database, complete variable analysis and aggregation through standardized and derived variables, and train a first prediction model through variable screening and parameter traversal; A model optimization module, configured to adjust the data in the business experience database based on business update data, optimize the parameters of the first prediction model, and obtain a second prediction model; A result generating module is used to generate demand data within a predetermined time period using the second prediction model.

9. An electronic device, characterized in that: It includes a processor and a memory, wherein the memory stores program code, and when the program code is executed by the processor, the processor executes any one of the business feedback-based demand forecasting methods in claims 1 to 6.

10. A computer-readable storage medium, characterized in that: It includes program code. When the storage medium is run on an electronic device, the program code is used to enable the electronic device to execute any one of the demand prediction methods based on business feedback in claims 1 to 7.

Images