CN107958297B - Product demand prediction method and product demand prediction device - Google Patents
Product demand prediction method and product demand prediction device Download PDFInfo
- Publication number
- CN107958297B CN107958297B CN201610902912.7A CN201610902912A CN107958297B CN 107958297 B CN107958297 B CN 107958297B CN 201610902912 A CN201610902912 A CN 201610902912A CN 107958297 B CN107958297 B CN 107958297B
- Authority
- CN
- China
- Prior art keywords
- product
- current
- stage
- predicted value
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Entrepreneurship & Innovation (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Human Resources & Organizations (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a product demand prediction method, which comprises the following steps: sampling a product adjustment record according to the product stage and the adjustment state to obtain a first sample set, and establishing a judgment model according to the first sample set; sampling the historical predicted value of the product according to the initial predicted value and the corrected predicted value to obtain a second sample set, and establishing a target regression model according to the second sample set; acquiring an initial predicted value of a current stage and a current product stage; processing the current product stage by using a judgment model to determine the current adjustment state; and if the current adjustment state is in need of adjustment, substituting the initial predicted value of the current stage into the target regression model for calculation, and outputting the corrected predicted value of the current stage. The invention also provides a product demand forecasting device capable of realizing the method. The method can optimize the existing prediction result and improve the accuracy of product demand prediction.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a product demand prediction method and a product demand prediction device.
Background
Product demand forecasting is a key link in enterprise operation and is used for guiding production and stock of enterprises. Excessive product demand forecasts can result in excessive inventory and increased inventory cost risks. Too little demand forecast may result in a low order fulfillment rate and reduced customer satisfaction. With the development of data science and the increasing importance of enterprises on data, many enterprises begin to use technologies related to big data for product demand forecasting. The product demand prediction method comprises a time sequence prediction method, a multiple regression analysis method and the like, and the demand prediction Model comprises an Autoregressive Integrated Moving Average Model (ARIMA for short), a Ridge regression Model (also called Ridge regression Model) and the like.
In the prior art, the product demand prediction method is roughly as follows: establishing a time series model according to historical data (such as historical product orders) of product demands; establishing a regression model according to a forecasting factor (such as seasonality, weather, forecasting factor correlation coefficient, promotion activity influence and the like) of the product demand; evaluating the time series model and the regression model respectively through a simulator; according to the evaluation result, the time series model and the regression model are combined by using a Bayesian integration method, and a prediction result is output, as shown in FIG. 1.
The existing demand forecasting method relies on a known regular forecasting factor, and effective estimation can be carried out only when the known forecasting factor appears more than twice. However, the existing demand forecasting model has limited considered forecasting factors, and many implicit forecasting factors are not added into the demand forecasting model, so that the implicit forecasting factors have great influence on the demand in practical application, so that the forecasting result and the actual demand are often in great difference, and the forecasting accuracy is not high.
Disclosure of Invention
The invention provides a product demand forecasting method and a product demand forecasting device, which can improve the forecasting accuracy.
A first aspect provides a product demand forecasting method, including: the product demand forecasting device samples the product adjustment record according to the first sample attribute to obtain a first sample set, and a judgment model is established according to the first sample set; sampling the product history predicted value according to the second sample attribute to obtain a second sample set, and establishing a target regression model according to the second sample set; acquiring data to be processed, wherein the data to be processed comprises an initial predicted value of a current stage and a current product stage; processing the current product stage by using a judgment model to determine the current adjustment state; if the current adjustment state is in need of adjustment, substituting the initial predicted value of the current stage into a target regression model for calculation, and outputting a corrected predicted value of the current stage, wherein the corrected predicted value of the current stage is used for making a production plan of the current stage; and if the current adjustment state is not required to be adjusted, outputting the initial predicted value of the current stage. The first sample attribute comprises a product stage and an adjustment state, the second sample attribute comprises an initial predicted value and a correction predicted value, and the product stage is one stage in the life cycle of the product.
Therefore, the predicted value is corrected to be in a stage that the service expert combines with the product, the expert experience is trained to obtain the improved demand prediction model, and compared with the existing product demand prediction system, the improved demand prediction model in the embodiment of the invention has more prediction factors considered, so that the product demand prediction method provided by the invention can improve the prediction precision.
In a possible implementation manner, the product demand prediction apparatus may specifically establish the judgment model according to the first sample set by the following steps: the product demand prediction device selects a first training set from the first sample set, trains the first training set to obtain a judgment model, and the judgment model is a logistic regression model, a decision tree model or a random forest model. According to the embodiment of the invention, all or part of the samples in the first sample set can be selected to establish various judgment models, and then the judgment models are utilized to predict the product requirements, so that the embodiment of the invention has good flexibility.
In another possible implementation manner, the product demand prediction device may select a first test set from the first sample set, test the judgment model by using the first test set to obtain a first test result, and compare the first test result with the historical data of the adjustment state to obtain a first prediction accuracy; and when the first prediction accuracy is not less than a first preset threshold, executing a step of processing the current product stage by using a judgment model and determining the current adjustment state. The embodiment of the invention can test the judgment model, and when the judgment model has good accuracy, the judgment model is used for prediction, thereby ensuring the accuracy of the product demand prediction result.
Further, in another possible implementation manner, the product demand prediction apparatus processes the current product stage using the judgment model, and the determining of the current adjustment state may be specifically implemented in the following manner: and the product demand prediction device processes the current product stage and the service expert information of the current stage by using the judgment model to determine the current adjustment state. The product adjustment record comprises historical data of the business expert information, the first sample attribute further comprises the business expert information, and the data to be processed comprises the business expert information of the current stage. The embodiment of the invention takes the service expert as an input parameter of the judgment model, can reflect the influence of the service expert on the product demand prediction method, and can further improve the prediction accuracy.
In another possible implementation manner, the product demand prediction apparatus processes the current product stage using the judgment model, and the determining of the current adjustment state may specifically be implemented by: the product demand forecasting device processes the current product stage and the product supply information of the current stage by using the judgment model, and determines the current adjustment state. The product adjustment record includes historical data of product supply information, the first sample attribute further includes product supply information, the data to be processed includes product supply information of the current stage, and the product supply information includes at least one of a delivery date, a supply location, or a number of suppliers. According to the embodiment of the invention, the product supply information is used as the input parameter of the judgment model, so that the influence of the product supply information on the product demand prediction method can be reflected, and the prediction accuracy can be further improved.
Further, in another possible implementation manner, the product demand prediction apparatus may specifically establish the target regression model according to the second sample set by the following steps: and the product demand prediction device selects a second training set from the second sample set, and trains the second training set to obtain a target regression model. The embodiment of the invention can select all or part of the samples in the second sample set to establish the target regression model.
Further, in another possible implementation manner, the product demand prediction apparatus may further select a second test set from the second sample set, test the target regression model by using the second test set to obtain a second test result, calculate an offset between the second test result and the historical order quantity, and determine a second prediction accuracy according to the offset; and when the second prediction accuracy is not less than a second preset threshold, the product demand prediction device carries out the step of substituting the initial prediction value of the current stage into the target regression model for calculation and outputting the corrected prediction value of the current stage. The embodiment of the invention can test the target regression model, and when the target regression model has good accuracy, the target regression model is used for prediction, so that the accuracy of the product demand prediction result is ensured.
In another possible implementation manner, after substituting the initial predicted value of the current stage into the target regression model for calculation and outputting the corrected predicted value of the current stage, the product demand prediction apparatus may further record the current product stage, the current adjustment state, the initial predicted value of the current stage, and the corrected predicted value of the current stage in the database.
A second aspect provides a product demand prediction apparatus capable of implementing the product demand prediction method of the first aspect. The function can be realized by hardware, and can also be realized by executing corresponding software by hardware.
As can be seen from the above embodiments, the present invention has the following advantages:
the product demand forecasting device can sample the product adjustment record to obtain a first sample set, and a judgment model is established according to the first sample set; sampling the historical predicted value of the product to obtain a second sample set, and establishing a target regression model according to the second sample set; processing the current product stage by using a judgment model to determine the current adjustment state; if the current adjustment state is in need of adjustment, substituting the initial predicted value of the current stage into the target regression model for calculation, and outputting the corrected predicted value of the current stage; and if the current adjustment state is not required to be adjusted, outputting the initial predicted value of the current stage. According to the method, the judgment model and the target regression model can be established through irregular relevant information (such as product stages) of the product, and then the existing prediction result is optimized by utilizing the judgment model and the target regression model, so that the prediction result is closer to the actual product demand, and the accuracy of product demand prediction is improved.
Drawings
FIG. 1 is a schematic diagram of a product demand forecasting method in the prior art;
FIG. 2 is a flow chart illustrating a product demand forecasting method according to an embodiment of the present invention;
FIG. 3 is a schematic flow chart of a product demand forecasting method according to an embodiment of the present invention;
FIG. 4 is a schematic flow chart of a product demand forecasting method according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a product demand forecasting apparatus according to an embodiment of the present invention;
FIG. 6 is another schematic diagram of a product demand forecasting apparatus in accordance with an embodiment of the present invention;
FIG. 7 is another schematic diagram of a product demand forecasting apparatus in accordance with an embodiment of the present invention;
FIG. 8 is another schematic diagram of a product demand forecasting apparatus in accordance with an embodiment of the present invention;
fig. 9 is a schematic diagram of a user equipment in an embodiment of the present invention.
Detailed Description
The following description will be given of an application environment of the product demand prediction method provided by the present invention, and the product demand prediction method of the present invention may be applied to both user equipment and a data processing system including the user equipment and a server.
The user equipment may be any electronic equipment such as a mobile phone, a tablet computer, a Personal Digital Assistant (PDA), a Point of Sales (POS), or a vehicle-mounted computer.
A server refers to a device that provides computing services. In a network environment, servers can be classified into a file server, a database server, an application server, a WEB server, and the like according to the type of service provided by the server.
According to the product demand prediction method in the prior art, a prediction factor which has a great influence on a prediction result is often ignored, so that the prediction result and an actual demand have great difference, and the prediction accuracy is not high. In practical application, the service expert also needs to adjust the initial prediction result obtained in the prior art by combining the influence factors of the current stage. In order to solve the above problems, the present invention provides a product demand prediction method, which has a core idea of obtaining a history record of irregular product-related information, such as a history prediction result and a product adjustment record, respectively training the history prediction result and the product adjustment record to obtain a regression model and a judgment model, and then optimizing the existing prediction result by using the judgment model and the regression model, so as to obtain a more accurate prediction result.
The product demand forecasting method provided by the invention is described below by taking user equipment as a product demand forecasting device as an example. Referring to fig. 2, an embodiment of a product demand prediction method according to the present invention includes:
step 201, the user equipment samples the product adjustment record according to the first sample attribute to obtain a first sample set, and establishes a judgment model according to the first sample set.
In this embodiment, the attributes of the product adjustment record may include a product stage and an adjustment state, and may further include other product-related information, such as service expert information (i.e., planner information), a product model, a product type, an initial predicted value or a corrected predicted value, and the like, which is not limited herein. The time for selecting the product adjustment record can be set according to actual conditions, such as 30 days, 10 weeks, 12 months, and the like, and is not limited herein. The product life cycle can be generally divided into a research and development period, a small batch release period, formal production and a product end period, wherein the product phase refers to one phase of the product life cycle. For example, when a new product with the same function is successfully developed, the product stage of the original product can be modified from the regular production to the final stage of the product. The first sample attribute may include a product stage and an adjustment state, and the user equipment samples the product adjustment record according to the first sample attribute to obtain a first sample set, and establishes the judgment model according to the first sample set.
The establishing, by the user equipment, the judgment model according to the first sample set may specifically be: the user equipment can select a plurality of samples from the first sample set as a first training set, and then train the first training set to obtain the judgment model. The method for training the first training set by the user equipment to obtain the judgment model can be specifically realized in the following way:
in an optional embodiment, the user equipment trains the first training set to obtain a logistic regression model。The method comprises the following specific steps: (1) constructing a prediction function h (x)i)＝P(y＝1|xi(ii) a θ); (2) constructing a loss function J (theta); (3) and calculating a regression parameter theta of the logistic regression model according to a gradient descent method. Wherein x isiFor the ith in the first training setAnd (4) sampling. P (y ═ 1| x)i(ii) a Theta) represents the probability that an adjustment is required when the product phase is the end of the product. The Logistic regression model may be a Logistic function (also known as sigmoid function).
In another alternative embodiment, the user equipment may train the first training set to obtain the decision tree model. In another alternative embodiment, the user equipment may train the first training set to obtain a random forest model. Specifically, the training algorithm of the decision tree model or the random forest model may refer to a decision tree algorithm or a random forest algorithm in the prior art, which is not described herein again.
Step 202, the user equipment samples the product history predicted value according to the second sample attribute to obtain a second sample set, and a target regression model is established according to the second sample set.
The product history predicted value comprises the history data of the initial predicted value and the history data of the corrected predicted value, and can also comprise other product related information, such as an adjustment amount and the like. The initial predicted value is a predicted result obtained by calculation by using the existing demand prediction method, the corrected predicted value is a predicted result determined by a service expert, and the adjustment quantity is a difference value between the corrected predicted value and the initial predicted value and is used for representing the adjustment quantity of the product predicted value.
The second sample attribute comprises an initial predicted value and a corrected predicted value, the user equipment samples the product history predicted value according to the second sample attribute to obtain a second sample set, and a target regression model is established according to the second sample set.
The establishing, by the user equipment, the target regression model according to the second sample set may specifically be: the user equipment may select a plurality of samples from the second sample set as a second training set, and train the second training set to obtain a target regression model, where the target regression model is a regression model corresponding to the second training set. The target regression model may be a linear regression model or a nonlinear regression model.
In an optional embodiment, the training of the second training set to obtain the target regression model may specifically be: and constructing a target regression model as Z ═ d × K. Wherein K is the initial predicted value, Z is the corrected predicted value, and d is the regression parameter of the target regression model. After the value of d is calculated from the second training set, a target regression model is established.
Specifically, the user equipment may obtain an initial predicted value of the current stage according to historical data of product requirements, and may also obtain the current product stage. The historical data for the product demand may be a product historical order volume. It should be noted that step 201 to step 203 are not in a fixed sequence. The process of the user equipment obtaining the initial predicted value of the current stage and the process of the user equipment obtaining the current product stage have no fixed sequence, which is not limited herein.
And step 204, the user equipment processes the current product stage by using the judgment model to determine the current adjustment state.
Specifically, after the current product stage and the judgment model are obtained, the user equipment takes the current product stage as an input parameter of the judgment model, and calculates the current adjustment state.
For example, if the current product stage is the final stage of the product, and the determination model is the logistic regression model, the calculation result is 1, which indicates that the current adjustment state is required to be adjusted, step 205 is executed, and if the current product stage is normal production, the calculation result is 0, which indicates that the current adjustment state is not required to be adjusted, step 206 is executed.
And step 205, if the current adjustment state is in need of adjustment, substituting the initial predicted value of the current stage into the target regression model for calculation, and outputting the corrected predicted value of the current stage.
And if the current adjustment state is in need of adjustment, the user equipment takes the initial predicted value of the current stage as an input parameter of the target regression model, and calculates and outputs the corrected predicted value of the current stage. The user equipment can display the correction predicted value, and the user can make a production plan of the current stage according to the correction predicted value of the current stage. It can be understood that the user equipment may output the demand prediction table according to the initial predicted value of the current stage, the corrected predicted value of the current stage, the adjustment amount, the product related information such as the service expert, and the like.
And step 206, if the current adjustment state is not required to be adjusted, outputting the initial predicted value of the current stage.
If the current adjustment state is not needed to be adjusted, the user equipment can output and display the initial predicted value of the current stage, and at the moment, the user can make a production plan of the current stage according to the initial predicted value of the current stage.
In practical application, because some prediction factors have no regularity, for example, the product stage changes state due to the successful development of products with the same function, the existing demand prediction model is difficult to accurately predict the demand of products. In the case that the product life cycle enters the end of the product, the actual product demand will drop, and the initial prediction value is greatly different from the actual product demand because the initial prediction value does not take into account the state change of irregular prediction factors (such as the product stage).
According to the method, the influence of the prediction factors on the product requirements can be estimated through the product adjustment records, the historical prediction results and the product adjustment records are trained to obtain the regression model and the judgment model, and then the regression model and the judgment model are used for optimizing the existing prediction results, so that the prediction results are more in line with the actual requirements, and the prediction accuracy is improved.
Based on the embodiment shown in fig. 2, in an optional embodiment of the present invention, after step 201 and before step 204, the product demand prediction method further includes:
the user equipment selects a first test set from the first sample set, tests the judgment model by using the first test set to obtain a first test result, and compares the first test result with historical data of the adjustment state to obtain a first prediction accuracy rate; when the first prediction accuracy is not less than the first preset threshold, step 204 is executed.
In this embodiment, the first preset threshold is used to measure the prediction accuracy of the judgment model. The user equipment may determine a prediction accuracy of the model based on the sample tests in the first test set. If the prediction accuracy is not smaller than the first preset threshold, the judgment model is reliable and has reasonable reliability. If the prediction accuracy is smaller than a first preset threshold, it indicates that the judgment model is not reliable enough, and the reliability is low, the current adjustment state may not be determined by using the judgment model, and at this time, the user equipment may continue to search for an appropriate regression parameter through the first training set to obtain a better judgment model. Generally, the samples in the first test set are not overlapped with the samples in the first training set, so as to ensure the accuracy of the test result.
For example, the first test set includes n1 samples, and the first test result includes n1 adjustment states, where m1 adjustment states are consistent with the historical adjustment states, and the prediction accuracy is m1/n 1. The first preset threshold is 98% as an example, if m1/n1 is greater than or equal to 98%, the judgment model is determined to be credible, and if m1/n1 is less than 98%, the judgment model is determined to be unreliable.
Based on the embodiment shown in fig. 2, in another alternative embodiment of the present invention, after step 202 and before step 205, the product demand forecasting method further comprises:
the user equipment selects a second test set from the second sample set, tests the target regression model by using the second test set to obtain a second test result, calculates the offset between the second test result and the historical order quantity, and determines a second prediction accuracy according to the offset; when the second prediction accuracy is not less than the second preset threshold, step 205 is executed.
In this embodiment, the second preset threshold is used to measure the accuracy of the regression model corresponding to the second training set. The user equipment may obtain the prediction accuracy of the regression model from the samples in the second test set. If the prediction accuracy is not less than the second preset threshold, the logistic regression model is reliable and has reasonable accuracy. If the prediction accuracy is smaller than a second preset threshold, which indicates that the regression model is not accurate enough and has low reliability, the user equipment does not use the regression model to calculate the corrected predicted value of the current stage, and at this time, the user equipment can continue to search for appropriate regression parameters through a second training set to obtain a better regression model.
Specifically, if the offset is smaller than the preset offset threshold, it may be determined that the obtained correction prediction value is accurate. And otherwise, determining that the obtained correction predicted value is not accurate enough, and calculating by using the regression model to obtain a correction predicted value with a larger error with the actual order quantity. Generally, the samples in the second test set do not overlap with the samples in the second training set, so as to ensure the accuracy of the test result.
For example, the second test set includes n2 samples, the second test result includes n2 correction predicted values, where the difference between the m2 correction predicted values and the historical order amount is smaller than the preset offset, and the prediction accuracy is m2/n 2. The second preset threshold is 95% as an example, if m2/n2 is greater than or equal to 95%, the judgment model is determined to be credible, and if m2/n2 is less than 95%, the judgment model is determined to be unreliable, and at this time, the user equipment can continue to search for appropriate regression parameters through the second training set to obtain a better regression model.
Based on the embodiment shown in fig. 2, in another alternative embodiment of the present invention, after step 205, the product demand forecasting method further includes:
and the user equipment records the current product stage, the current adjustment state, the initial predicted value of the current stage and the corrected predicted value of the current stage in a database.
In this embodiment, after the user equipment obtains the current product stage, the current adjustment state, the initial predicted value of the current stage, and the corrected predicted value of the current stage, the data may be recorded in the database as historical data of product demand prediction of the next stage.
In order to further improve the prediction accuracy of the product demand prediction model, the method can also take other attributes of the product as prediction factors to construct a better product demand prediction model. Referring to fig. 3, another embodiment of the product demand forecasting method according to the present invention includes:
step 301, the user equipment samples the product adjustment record according to the product stage, the service expert information and the adjustment state to obtain a first sample set, and establishes a judgment model according to the first sample set.
In this embodiment, the attributes of the product adjustment record at least include a product phase, service expert information, and an adjustment state, and the first sample attribute includes a product phase, service expert information, and an adjustment state. The user device samples from the product adjustment record according to the first sample attribute to obtain a first sample set. The user equipment can select a plurality of samples from the first sample set as a first training set, and then train the first training set to obtain the judgment model. The prediction factors of the judgment model in this embodiment include product stage and business expert information.
And step 302, the user equipment samples the product history predicted value according to the second sample attribute to obtain a second sample set, and establishes a target regression model according to the second sample set.
And step 304, the user equipment processes the current product stage and the current stage of the service expert information by using the judgment model to determine the current adjustment state.
Specifically, after acquiring the current product stage and the current stage of the service expert information, the user equipment may use the current product stage and the current stage of the service expert information as input parameters of the judgment model to calculate the current adjustment state. The current stage refers to a period corresponding to the current product stage.
And 305, substituting the initial predicted value of the current stage into the target regression model for calculation if the current adjustment state is in need of adjustment, and outputting the corrected predicted value of the current stage.
And step 306, outputting the initial predicted value of the current stage if the current adjustment state is not required to be adjusted.
In this embodiment, step 302 is similar to step 202, step 305 is similar to step 205, and step 306 is similar to step 206, which are not repeated herein.
As can be seen from the above, in the embodiment, the service expert information is used as the prediction factor of the judgment model to represent the influence of the service expert information on the adjustment state, so that the judgment result of the judgment model is optimized, and the accuracy of product demand prediction is further improved.
The product demand prediction method provided by the present invention is further described below with product supply information as a predictor. Referring to fig. 4, another embodiment of the product demand forecasting method according to the present invention includes:
In this embodiment, the attributes of the product adjustment record at least include a product stage, product supply information, and an adjustment status, and the product supply information includes at least one of a delivery date, a supply location, or a number of suppliers. The first sample attributes include product phase, product supply information, and adjustment status. The user device samples from the product adjustment record according to the first sample attribute to obtain a first sample set. The user equipment can select a plurality of samples from the first sample set as a first training set, and then train the first training set to obtain the judgment model. The prediction factors of the judgment model in this embodiment include product stage and product supply information.
And step 402, the user equipment samples the product history predicted value according to the second sample attribute to obtain a second sample set, and establishes a target regression model according to the second sample set.
Specifically, after obtaining the product supply information of the current product stage and the current stage, the user equipment may use the product supply information of the current product stage and the current stage as input parameters of the determination model to calculate the current adjustment state.
And 405, if the current adjustment state is in need of adjustment, substituting the initial predicted value of the current stage into the target regression model for calculation, and outputting the corrected predicted value of the current stage.
And step 406, outputting the initial predicted value of the current stage if the current adjustment state is not required to be adjusted.
In this embodiment, step 402 is similar to step 202, step 405 is similar to step 205, and step 406 is similar to step 206, which are not repeated herein.
As can be seen from the above, the influence of the change of the product supply information (such as the shelf life, the supply location, the number of suppliers, and the like) on the product demand is significant, and in the embodiment, the product supply information is used as the prediction factor of the judgment model to represent the influence of the product supply information on the adjustment state, so that the judgment result of the judgment model can be optimized, and the accuracy of product demand prediction is further improved.
It can be understood that the invention can also use the combination of business experts, product supply information and/or other product related information as the predictor of the product demand prediction method, and then use the predictor to establish the judgment model.
For convenience of understanding, the product demand prediction method provided by the embodiment of the present invention is described in detail in a specific application scenario as follows:
the computer retrieves the product adjustment record from the database as shown in table 1. When the adjustment state is "yes", it indicates that adjustment is necessary, and when the adjustment state is "no", it indicates that adjustment is not necessary.
| Numbering | Month of the year | Business expert | Product stage | Adjusting the state |
| 1 | 1 month | Expert armor | Formal production | Whether or not |
| 2 | 2 month | Expert armor | End of product | Is that |
| 3 | 3 month | Expert B | Formal production | Whether or not |
| … | … | … | … | … |
| 10 | 10 month | Expert B | End of product | Is that |
TABLE 1
When the first sample attribute includes the product stage and the adjustment status, the training set 1 obtained by the computer selecting 9 samples is shown in table 2.
| Numbering | Product stage | Adjusting the state |
| 1 | Formal production | Whether or not |
| 2 | End of product | Is that |
| 3 | Formal production | Whether or not |
| … | … | … |
| 9 | End of product | Is that |
TABLE 2
In the training set 1, the probability that the product phase is the end of production is 50%, the probability that the product phase is normal production is 50%, the probability that the adjustment state is "yes" under the condition of the end of production is 100%, and the probability that the adjustment state is "yes" under the condition of normal production is 0. Then the logistic regression function obtained from training set 2The number may be: p (y ═ 1| x)i；θ)＝g(0.5+0.1x1-0.1x2)，x1Indicates end of product, x2Indicating normal production.
The product history prediction values obtained from the database by the computer are shown in table 3.
TABLE 3
When the second sample attribute is the initial predicted value and the corrected predicted value, the training set 2 obtained by the computer selecting 8 samples is shown in table 4.
| Numbering | Initial prediction value | Correcting predicted values |
| 1 | 3280 | 1000 |
| 2 | 948 | 200 |
| … | … | … |
| 8 | 2429 | 1400 |
TABLE 4
When the regression function is trained according to training set 2: when Z equals d × K, d equals 0.53.
The current stage takes 11 months as an example, and the computer obtains an initial predicted value of 11 months and the current product stage. Taking the final stage of the product as an example at the current product stage, taking the final stage of the product as an input parameter (namely x) of the judgment model1＝1，x20), the current adjustment state P is calculated (y 1| x)1(ii) a θ) is g (0.6) is 1, indicating that the current adjustment state is adjustment-required. The initial predicted value in month 11 is 2800 as an example, and the computer calculates and displays a corrected predicted value Z1484 using 2800 as an input parameter of Z0.53 × K.
If the current product stage is normal production, the computer takes the normal production as the input parameter (i.e. x) of the judgment model2＝1，x10), the current adjustment state P is calculated (y 1| x)2(ii) a θ) is g (0.4) is 0, which indicates that the current adjustment state is not required to be adjusted, and outputs and displays the prediction result K as 2800. Under the condition that the life cycle of the product enters the final stage of the product, the actual product demand is reduced, and the initial predicted value does not consider the change of the product stage, so that the difference between the initial predicted value and the actual product demand is larger.
Referring to fig. 5, an embodiment of a product demand forecasting apparatus 500 according to the present invention includes:
the first training module 501 is configured to sample a product adjustment record according to a first sample attribute to obtain a first sample set, and establish a judgment model according to the first sample set, where the first sample attribute includes a product stage and an adjustment state;
the second training module 502 is configured to sample the product history predicted value according to a second sample attribute to obtain a second sample set, and establish a target regression model according to the second sample set, where the second sample attribute includes an initial predicted value and a corrected predicted value;
the data acquiring module 503 is configured to acquire data to be processed, where the data to be processed includes an initial predicted value of a current stage and a current product stage;
a state determination module 504, configured to process the current product stage using the judgment model, and determine a current adjustment state;
the prediction module 505 is configured to substitute the initial predicted value of the current stage into the target regression model for calculation if the current adjustment state is in need of adjustment, and output a corrected predicted value of the current stage, where the corrected predicted value of the current stage is used for making a production plan of the current stage; and if the current adjustment state is not required to be adjusted, outputting the initial predicted value of the current stage.
Based on the embodiment shown in fig. 5, in an optional embodiment of the present invention, the first training module 501 is specifically configured to select a first training set from the first sample set, train the first training set to obtain a judgment model, where the judgment model is a logistic regression model, a decision tree model, or a random forest model.
Based on the embodiment or the alternative embodiment shown in fig. 5, in another alternative embodiment of the present invention, the product demand forecasting device 500 further includes:
the first testing module 601 is configured to select a first testing set from the first sample set, test the judgment model by using the first testing set to obtain a first testing result, and compare the first testing result with historical data of the adjustment state to obtain a first prediction accuracy; when the first prediction accuracy is not less than the first preset threshold, the determination state module 504 is triggered to execute the step of determining the current adjustment state by processing the current product stage using the judgment model.
Based on the embodiment or the alternative embodiment shown in fig. 5, in another alternative embodiment of the present invention, the product adjustment record includes historical data of the business expert information, the first sample attribute further includes the business expert information, and the data to be processed includes the business expert information of the current stage. The state determining module 504 is specifically configured to process the current product stage and the current stage of the service expert information using the determination model, and determine a current adjustment state.
Based on the embodiment or alternative embodiment shown in fig. 5, in another alternative embodiment of the present invention, the product adjustment record includes historical data of product supply information, the first sample attribute further includes product supply information, the data to be processed includes product supply information of the current stage, and the product supply information includes at least one of a period of time, a supply location, or a number of suppliers;
the state determining module 504 is specifically configured to process the current product stage and the product supply information of the current stage by using the determination model, and determine a current adjustment state.
Based on the embodiment or the alternative embodiment shown in fig. 5, in another alternative embodiment of the present invention, the product demand forecasting device 500 further includes:
the second testing module 701 is configured to select a second testing set from the second sample set, test the target regression model by using the second testing set to obtain a second testing result, calculate an offset between the second testing result and the historical order quantity, and determine a second prediction accuracy according to the offset;
and when the second prediction accuracy is not less than a second preset threshold, triggering the prediction module 505 to perform the step of substituting the initial prediction value of the current stage into the target regression model for calculation, and outputting the corrected prediction value of the current stage.
Based on the embodiment or the alternative embodiment shown in fig. 5, in another alternative embodiment of the present invention, the product demand forecasting device 500 further includes:
a recording module 801, configured to record the current product stage, the current adjustment state, the initial predicted value of the current stage, and the corrected predicted value of the current stage in a database.
For convenience of understanding, the following describes in detail the interaction between modules in the product demand forecasting apparatus provided by the embodiment of the present invention in a specific application scenario:
the first training module 501 retrieves the product adjustment record from the database as shown in table 1. When the adjustment state is "yes", it indicates that adjustment is necessary, and when the adjustment state is "no", it indicates that adjustment is not necessary.
| Numbering | Month of the year | Business expert | Product stage | Adjusting the state |
| 1 | 1 month | Expert armor | Formal production | Whether or not |
| 2 | 2 month | Expert armor | End of product | Is that |
| 3 | 3 month | Expert B | Formal production | Whether or not |
| … | … | … | … | … |
| 10 | 10 month | Expert B | End of product | Is that |
TABLE 1
When the first sample attribute includes the product stage and the adjustment state, the training set 1 obtained by the first training module 501 selecting 9 samples is shown in table 2.
TABLE 2
In the training set 1, the probability that the product phase is the end of production is 50%, the probability that the product phase is normal production is 50%, the probability that the adjustment state is "yes" under the condition of the end of production is 100%, and the probability that the adjustment state is "yes" under the condition of normal production is 0. The logistic regression function obtained by the first training module 501 from the training set 2 may be: p (y ═ 1| x)i；θ)＝g(0.5+0.1x1-0.1x2)，x1Indicates end of product, x2Indicating normal production.
The product history prediction values obtained by the second training module 502 are shown in table 3.
| Numbering | Initial prediction value | Correcting predicted values | Amount of adjustment |
| 1 | 14474 | 14474 | 0 |
| 2 | 3280 | 1000 | 2280 |
| 3 | 948 | 200 | 748 |
| … | … | … | … |
| 10 | 2429 | 1400 | 1029 |
TABLE 3
When the second sample attribute is the initial prediction value and the corrected prediction value, the second training module 502 selects 8 samples to obtain a training set 2 as shown in table 4.
| Numbering | Initial prediction value | Correcting predicted values |
| 1 | 3280 | 1000 |
| 2 | 948 | 200 |
| … | … | … |
| 8 | 2429 | 1400 |
TABLE 4
When the regression function is trained to Z ═ d × K according to training set 2, d ═ 0.53 is calculated.
Taking 11 months as an example in the current stage, the data obtaining module 503 obtains an initial predicted value and a current product stage of the current stage respectively. Taking the final stage of the product as an example at the current product stage, taking the final stage of the product as an input parameter (namely x) of the judgment model1＝1，x20), the current adjustment state P is calculated (y 1| x)1(ii) a θ) ═ g (0.6) ═ 1, and indicates the current keyThe integral state is required to be adjusted. The initial predicted value in month 11 is 2800 as an example, and the prediction module 505 calculates and displays a corrected predicted value Z1484 using 2800 as an input parameter of Z0.53 × K.
If the current product stage is normal production, the determine state module 504 takes normal production as an input parameter (i.e., x) of the determination model2＝1，x10), the current adjustment state P is calculated (y 1| x)2(ii) a θ) — g (0.4) — 0, which indicates that the current adjustment state is not required to be adjusted, and the prediction module 505 outputs and displays the prediction result K ═ 2800.
In the following, the user equipment 900 in the embodiment of the present invention is introduced from the perspective of a hardware device, and the product demand prediction apparatus 500 in the embodiments shown in fig. 5 to 8 can all implement the product demand prediction method provided by the present invention based on the structure of the user equipment 900. One embodiment of the user equipment 900 provided by the present invention comprises:
the user equipment 900 comprises an input device 901, an output device 902, a processor 903, a memory 904, and a bus; the input device 901, the output device 902, the processor 903, and the memory 904 are connected to each other via a bus.
The Processor 903 may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the device may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA), or other Programmable logic devices.
The Memory 904 may include a Random Access Memory (RAM) implementation, and may also include a Non-Volatile Memory (NVM), such as at least one disk Memory.
The memory 904 is used for storing data and operation instructions, and the processor 903 may execute the product demand prediction processing method in the embodiments or optional embodiments shown in fig. 2 to 4 by calling the operation instructions stored in the memory 904.
In the embodiments provided in the present invention, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the above-described division of units is only one type of division of logical functions, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit may be stored in a computer-readable storage medium if it is implemented in the form of a software functional unit and sold or used as a separate product. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the above method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (16)
1. A method for predicting demand for a product, comprising:
sampling a product adjustment record according to a first sample attribute to obtain a first sample set, and establishing a judgment model according to the first sample set, wherein the first sample attribute comprises a product stage and an adjustment state;
sampling the product history predicted value according to a second sample attribute to obtain a second sample set, and establishing a target regression model according to the second sample set, wherein the second sample attribute comprises an initial predicted value and a corrected predicted value;
acquiring data to be processed, wherein the data to be processed comprises an initial predicted value of a current stage and a current product stage;
processing the current product stage by using the judgment model to determine a current adjustment state;
if the current adjustment state is in need of adjustment, substituting the initial predicted value of the current stage into the target regression model for calculation, and outputting a corrected predicted value of the current stage, wherein the corrected predicted value of the current stage is used for making a production plan of the current stage;
and if the current adjustment state is not required to be adjusted, outputting the initial predicted value of the current stage.
2. The method of claim 1, wherein the building a decision model from the first set of samples comprises:
and selecting a first training set from the first sample set, and training the first training set to obtain a judgment model, wherein the judgment model is a logistic regression model, a decision tree model or a random forest model.
3. The method of claim 2, wherein prior to said determining a current adjustment status using said decision model to process said current product stage, said method further comprises:
selecting a first test set from the first sample set, testing the judgment model by using the first test set to obtain a first test result, and comparing the first test result with historical data of an adjustment state to obtain a first prediction accuracy;
and when the first prediction accuracy is not smaller than a first preset threshold, executing the step of processing the current product stage by using the judgment model to determine the current adjustment state.
4. The method of claim 2, wherein the product adjustment record comprises historical data of business expert information, the first sample attribute further comprises business expert information, and the data to be processed comprises business expert information of a current stage;
the processing the current product stage using the judgment model, and determining a current adjustment state includes:
and processing the service expert information of the current product stage and the current stage by using the judgment model, and determining the current adjustment state.
5. The method of claim 2, wherein the product adjustment record includes historical data of product supply information, the first sample attribute further includes product supply information, the pending data includes product supply information for a current stage, the product supply information includes at least one of a date of shipment, a location of supply, or a quantity of suppliers;
the processing the current product stage using the judgment model, and determining a current adjustment state includes:
and processing the current product stage and the product supply information of the current stage by using the judgment model, and determining the current adjustment state.
6. The method of any one of claims 1 to 5, wherein the building a target regression model from the second sample set comprises:
and selecting a second training set from the second sample set, and training the second training set to obtain a target regression model.
7. The method of claim 6, wherein before the substituting the initial predicted value of the current stage into the target regression model calculation and outputting the corrected predicted value of the current stage, the method further comprises:
selecting a second test set from the second sample set, testing the target regression model by using the second test set to obtain a second test result, calculating the offset between the second test result and the historical order quantity, and determining a second prediction accuracy according to the offset;
and when the second prediction accuracy is not less than a second preset threshold, the step of substituting the initial predicted value of the current stage into the target regression model for calculation and outputting a corrected predicted value of the current stage is executed.
8. The method of claim 6, wherein after the substituting the initial predicted value of the current stage into the target regression model calculation and outputting the corrected predicted value of the current stage, the method further comprises:
and recording the current product stage, the current adjustment state, the initial predicted value of the current stage and the corrected predicted value of the current stage in a database.
9. A product demand forecasting apparatus comprising:
the first training module is used for sampling a product adjustment record according to a first sample attribute to obtain a first sample set, and establishing a judgment model according to the first sample set, wherein the first sample attribute comprises a product stage and an adjustment state;
the second training module is used for sampling the product history predicted value according to a second sample attribute to obtain a second sample set, and establishing a target regression model according to the second sample set, wherein the second sample attribute comprises an initial predicted value and a corrected predicted value;
the data acquisition module is used for acquiring data to be processed, wherein the data to be processed comprises an initial predicted value of a current stage and a current product stage;
a state determining module, configured to process the current product stage using the judgment model, and determine a current adjustment state;
the prediction module is used for substituting the initial predicted value of the current stage into the target regression model for calculation if the current adjustment state is in need of adjustment, and outputting a corrected predicted value of the current stage, wherein the corrected predicted value of the current stage is used for making a production plan of the current stage; and if the current adjustment state is not required to be adjusted, outputting the initial predicted value of the current stage.
10. The apparatus of claim 9, wherein the first training module is specifically configured to select a first training set from the first sample set, train the first training set to obtain a judgment model, and the judgment model is a logistic regression model, a decision tree model, or a random forest model.
11. The apparatus of claim 10, wherein the product demand forecasting device further comprises:
the first test module is used for selecting a first test set from the first sample set, testing the judgment model by using the first test set to obtain a first test result, and comparing the first test result with historical data of an adjustment state to obtain a first prediction accuracy;
and when the first prediction accuracy is not smaller than a first preset threshold, triggering the state determining module to execute the step of processing the current product stage by using the judgment model to determine the current adjustment state.
12. The apparatus of claim 10, wherein the product adjustment record comprises historical data of business expert information, the first sample attribute further comprises business expert information, and the data to be processed comprises business expert information of a current stage;
the state determining module is specifically configured to process the current product stage and the current stage of service expert information by using the judgment model, and determine a current adjustment state.
13. The apparatus of claim 10, wherein the product adjustment record comprises historical data of product supply information, the first sample attribute further comprises product supply information, the to-be-processed data comprises product supply information for a current stage, the product supply information comprises at least one of a date of delivery, a place of supply, or a number of suppliers;
the state determining module is specifically configured to process the current product stage and the product supply information of the current stage by using the determination model, and determine a current adjustment state.
14. The apparatus according to any one of claims 9 to 13, wherein the second training module is specifically configured to select a second training set from the second sample set, and train the second training set to obtain a target regression model.
15. The apparatus of claim 14, wherein the product demand forecasting device further comprises:
the second testing module is used for selecting a second testing set from the second sample set, testing the target regression model by using the second testing set to obtain a second testing result, calculating the offset of the second testing result and the historical order quantity, and determining a second prediction accuracy rate according to the offset;
and when the second prediction accuracy is not less than a second preset threshold, triggering the prediction module to carry out the step of substituting the initial prediction value of the current stage into the target regression model for calculation and outputting a corrected prediction value of the current stage.
16. The apparatus of claim 14, wherein the product demand forecasting device further comprises:
and the recording module is used for recording the current product stage, the current adjustment state, the initial predicted value of the current stage and the corrected predicted value of the current stage in a database.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610902912.7A CN107958297B (en) | 2016-10-17 | 2016-10-17 | Product demand prediction method and product demand prediction device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610902912.7A CN107958297B (en) | 2016-10-17 | 2016-10-17 | Product demand prediction method and product demand prediction device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107958297A CN107958297A (en) | 2018-04-24 |
| CN107958297B true CN107958297B (en) | 2021-10-26 |
Family
ID=61954388
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610902912.7A Active CN107958297B (en) | 2016-10-17 | 2016-10-17 | Product demand prediction method and product demand prediction device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107958297B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108829718B (en) * | 2018-05-07 | 2021-04-06 | 北京奇艺世纪科技有限公司 | Data processing method and device |
| CN109101612B (en) * | 2018-08-06 | 2022-06-17 | 优信数享（北京）信息技术有限公司 | Method, device and system for providing solution |
| CN109544076B (en) * | 2018-11-28 | 2021-06-18 | 北京百度网讯科技有限公司 | Method and apparatus for generating information |
| CN111717588B (en) * | 2019-03-20 | 2024-04-16 | 北京京东乾石科技有限公司 | Method, system, device and readable storage medium for managing auto-stereoscopic warehouse |
| CN110083802B (en) * | 2019-04-12 | 2022-05-03 | 重庆邮电大学 | Highway toll station vehicle flow big data prediction method based on multi-target regression |
| CN112668746A (en) * | 2019-10-15 | 2021-04-16 | 深圳怡化电脑股份有限公司 | Standby module demand prediction method and device, storage medium and equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819668A (en) * | 2010-04-27 | 2010-09-01 | 浙江大学 | Sales predicting model based on product intrinsic life cycle character |
| CN102142120A (en) * | 2010-02-01 | 2011-08-03 | 台湾积体电路制造股份有限公司 | Methods and systems for dynamic inventory control |
| CN102214338A (en) * | 2010-04-06 | 2011-10-12 | 上海驭策信息技术有限公司 | Sales forecasting system and method |
| CN104636823A (en) * | 2015-01-23 | 2015-05-20 | 中国农业大学 | Wind power prediction method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020165765A1 (en) * | 2001-05-03 | 2002-11-07 | Benny Sommerfeld | Method for managing a product strategy |
-
2016
- 2016-10-17 CN CN201610902912.7A patent/CN107958297B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102142120A (en) * | 2010-02-01 | 2011-08-03 | 台湾积体电路制造股份有限公司 | Methods and systems for dynamic inventory control |
| CN102214338A (en) * | 2010-04-06 | 2011-10-12 | 上海驭策信息技术有限公司 | Sales forecasting system and method |
| CN101819668A (en) * | 2010-04-27 | 2010-09-01 | 浙江大学 | Sales predicting model based on product intrinsic life cycle character |
| CN104636823A (en) * | 2015-01-23 | 2015-05-20 | 中国农业大学 | Wind power prediction method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107958297A (en) | 2018-04-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107958297B (en) | Product demand prediction method and product demand prediction device | |
| US11593817B2 (en) | Demand prediction method, demand prediction apparatus, and non-transitory computer-readable recording medium | |
| US8027897B2 (en) | System and method for optimizing financial performance generated by marketing investments under budget constraints | |
| CN108710905B (en) | Spare part quantity prediction method and system based on multi-model combination | |
| US8180693B2 (en) | Prediction of financial performance for a given portfolio of marketing investments | |
| CN110555578B (en) | Sales prediction method and device | |
| US20140207705A1 (en) | Residual risk analysis system, method and computer program product therefor | |
| CN112215546B (en) | Object page generation method, device, equipment and storage medium | |
| CN116188061B (en) | Commodity sales predicting method and device, electronic equipment and storage medium | |
| US20210065128A1 (en) | System and method for recruitment candidate equity modeling | |
| US11301763B2 (en) | Prediction model generation system, method, and program | |
| CN111652655A (en) | Commodity sales prediction method and device, electronic equipment and readable storage medium | |
| CN110866698A (en) | Device for assessing service score of service provider | |
| Mishra et al. | Failure prediction model for predictive maintenance | |
| JP2013131259A (en) | Integrated demand prediction device, integrated demand prediction method, and integrated demand prediction program | |
| CN114372735A (en) | Method and device for determining buffer stock quantity data in logistics supply chain | |
| US11586951B2 (en) | Evaluation system, evaluation method, and evaluation program for evaluating a result of optimization based on prediction | |
| JP2015032267A (en) | Demand prediction apparatus and program | |
| CN110689177A (en) | Method and device for predicting order preparation time, electronic equipment and storage medium | |
| US8180694B2 (en) | System and method for incorporating qualitative inputs into econometric models | |
| CN112132498A (en) | Inventory management method, device, equipment and storage medium | |
| KR102531633B1 (en) | Pharmaceutical industry technology asset evaluation system and method using technology asset contribution | |
| US20210304243A1 (en) | Optimization of markdown schedules for clearance items at physical retail stores | |
| CN115943399A (en) | Generation method, generation device, program, information processing method, and information processing device | |
| JP2021152693A (en) | Sales prediction system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |