CN111724211A - Offline store commodity sales prediction method, device and equipment - Google Patents
Offline store commodity sales prediction method, device and equipment Download PDFInfo
- Publication number
- CN111724211A CN111724211A CN202010614259.0A CN202010614259A CN111724211A CN 111724211 A CN111724211 A CN 111724211A CN 202010614259 A CN202010614259 A CN 202010614259A CN 111724211 A CN111724211 A CN 111724211A
- Authority
- CN
- China
- Prior art keywords
- sales data
- commodity
- sales
- data
- target
- 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.)
- Pending
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
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- 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
- G06Q30/0204—Market segmentation
- G06Q30/0205—Location or geographical consideration
Abstract
The application discloses a method, a device and equipment for predicting commodity sales of off-line stores, which are used for carrying out target feature extraction and hierarchical classification on first historical sales data of acquired commodities to obtain a plurality of training sets; respectively inputting each training set into a plurality of models to be trained for training to obtain a plurality of prediction models; respectively inputting first characteristics extracted based on second historical sales data of the target commodity into each prediction model to obtain a plurality of predicted sales volumes; calculating an error value of the actual sales volume of the target commodity and each predicted sales volume, and taking a prediction model corresponding to the minimum error value as an optimal prediction model; the third history sales data of the target commodity are obtained, the second characteristics extracted from the third history sales data of the target commodity are input into the optimal prediction model, and the sales volume of the target commodity in the preset time period is output, so that the technical problem of low prediction accuracy rate caused by the adoption of a single prediction model in the conventional sales volume prediction method is solved.
Description
Technical Field
The application relates to the technical field of data processing, in particular to a method, a device and equipment for predicting commodity sales of off-line stores.
Background
The existing commodity sales prediction methods adopt a single model for prediction, and can be mainly divided into two types, one type is a prediction model based on a time sequence, the prediction is carried out by processing the time sequence of a prediction target, and the model has weak learning capacity on non-sequence and non-linear relation data, so that the prediction accuracy is not high; the other type is a prediction model based on machine learning, historical data is trained through models such as a support vector machine, a random forest and a convolutional neural network, and when the difference between the data to be predicted and the historical data is large, the prediction accuracy rate is reduced. The commodity sales data of the off-line stores have different attributes such as regions and categories, so that the obtained commodity data are of multiple types, and the technical problem of low prediction accuracy exists when a single prediction model is adopted to predict the commodity data of multiple types.
Disclosure of Invention
The application provides a method, a device and equipment for predicting commodity sales of off-line stores, which are used for solving the technical problem of low prediction accuracy rate of the existing sales prediction method by adopting a single prediction model.
In view of the above, a first aspect of the present application provides a method for predicting commodity sales of an offline store, including:
acquiring first historical sales data of commodities of a plurality of offline stores;
performing target feature extraction and hierarchical classification on the first historical sales data to obtain a plurality of training sets;
respectively inputting each training set into a plurality of models to be trained for training to obtain a plurality of prediction models, wherein the number of the prediction models is the product of the number of the training sets and the number of the models to be trained, and the models to be trained at least comprise a machine learning model and a time sequence model;
performing target feature extraction on the acquired second historical sales data of the target commodity to obtain first features, and respectively inputting the first features into each prediction model to obtain a plurality of predicted sales volumes, wherein the time period of the second historical sales data from the current time is less than the time period of the first historical sales data from the current time;
calculating an error value of the actual sales volume of the target commodity and each predicted sales volume, and taking the prediction model corresponding to the minimum error value as an optimal prediction model;
and acquiring third historical sales data of the target commodity, extracting target features of the third historical sales data to obtain second features, inputting the second features into the optimal prediction model, and outputting the sales volume of the target commodity in a preset time period, wherein the time period of the third historical sales data from the current time is less than the time period of the second historical sales data from the current time.
Optionally, the performing target feature extraction on the first historical sales data further includes:
preprocessing the first historical sales data, wherein the preprocessing comprises data screening, data missing value filling or abnormal value modification.
Optionally, the target feature includes at least a seasonal feature, a holiday feature, a weather feature, or a store feature.
Optionally, the hierarchical classification is performed on the first historical sales data, and includes:
and hierarchically classifying the first historical sales data based on commodity attributes, wherein the commodity attributes at least comprise regions, commodity categories, sales volumes, volatility and sales frequency.
Optionally, the error value comprises a weighted average absolute percentage error value, a mean square error value, or a mean absolute error value.
A second aspect of the present application provides an offline store commodity sales amount prediction apparatus, including:
the system comprises a first acquisition unit, a second acquisition unit and a display unit, wherein the first acquisition unit is used for acquiring first historical sales data of commodities of a plurality of offline stores;
the classification unit is used for carrying out target feature extraction and hierarchical classification on the first historical sales data to obtain a plurality of training sets;
the training unit is used for inputting each training set into a plurality of models to be trained respectively for training to obtain a plurality of prediction models, the number of the prediction models is the product of the number of the training sets and the number of the models to be trained, and the models to be trained at least comprise a machine learning model and a time sequence model;
the extraction unit is used for extracting target characteristics of the obtained second historical sales data of the target commodity to obtain first characteristics, the first characteristics are respectively input into each prediction model to obtain a plurality of predicted sales volumes, and the time period of the second historical sales data from the current time is less than the time period of the first historical sales data from the current time;
a calculation unit, configured to calculate an error value between an actual sales volume of the target product and each predicted sales volume, and use the prediction model corresponding to the smallest error value as an optimal prediction model;
and the prediction unit is used for acquiring third historical sales data of the target commodity, extracting target characteristics of the third historical sales data to obtain second characteristics, inputting the second characteristics into the optimal prediction model, and outputting the sales volume of the target commodity in a preset time period, wherein the time period of the third historical sales data from the current time is less than the time period of the second historical sales data from the current time.
Optionally, the method further includes:
and the preprocessing unit is used for preprocessing the first historical sales data, and the preprocessing comprises data screening, data missing value filling or abnormal value modification.
Optionally, the target feature includes at least a seasonal feature, a holiday feature, a weather feature, or a store feature.
Optionally, the error value comprises a weighted average absolute percentage error value, a mean square error value, or a mean absolute error value.
A third aspect of the present application provides an offline store commodity sales predicting apparatus, comprising a processor and a memory;
the memory is used for storing program codes and transmitting the program codes to the processor;
the processor is used for executing any one of the offline store commodity sales amount prediction methods in a single aspect according to instructions in the program code.
According to the technical scheme, the method has the following advantages:
the application provides a method for predicting commodity sales of off-line stores, which comprises the following steps: acquiring first historical sales data of commodities of a plurality of offline stores; performing target feature extraction and hierarchical classification on the first historical sales data to obtain a plurality of training sets; respectively inputting each training set into a plurality of models to be trained for training to obtain a plurality of prediction models, wherein the number of the prediction models is the product of the number of the training sets and the number of the models to be trained, and the models to be trained at least comprise a machine learning model and a time sequence model; performing target feature extraction on the acquired second historical sales data of the target commodity to obtain first features, and respectively inputting the first features into each prediction model to obtain a plurality of predicted sales volumes, wherein the time period of the second historical sales data from the current time is less than the time period of the third historical sales data from the current time; calculating an error value of the actual sales volume of the target commodity and each predicted sales volume, and taking a prediction model corresponding to the minimum error value as an optimal prediction model; and acquiring third historical sales data of the target commodity, extracting target characteristics of the third historical sales data to obtain second characteristics, inputting the second characteristics into the optimal prediction model, and outputting the sales volume of the target commodity in a preset time period, wherein the time period of the third historical sales data from the current time is less than the time period of the second historical sales data from the current time.
According to the method for predicting the commodity sales volume of the off-line stores, first historical sales data of commodities of a plurality of off-line stores are obtained, target feature extraction and hierarchical classification are carried out on the first historical sales data to obtain a plurality of training sets, and through the hierarchical classification, the features of each training set are outstanding, each training set has more consistent features, and a model to be trained can be better trained; each data set is respectively input into a plurality of models to be trained for training, and a plurality of models to be trained such as a machine learning model and a time sequence model are used for training, so that the problem of low prediction accuracy of a single model is avoided; the method comprises the steps of extracting target characteristics of the obtained second historical sales data of the target commodity, inputting the second historical sales data into each prediction model to obtain a plurality of predicted sales, measuring the error values of the predicted sales and the actual sales back and forth by calculating, selecting the optimal prediction model suitable for the target commodity, predicting the sales of the target commodity in a preset time period by using the optimal prediction model, and measuring back by using the second historical sales data of the target commodity, so that the prediction model most suitable for the target commodity is selected, the prediction accuracy is improved, and the technical problem that the prediction accuracy is low due to the fact that a single prediction model is adopted in the existing sales prediction method is solved.
Drawings
Fig. 1 is a schematic flow chart of a method for predicting the commodity sales of an offline store according to an embodiment of the present disclosure;
fig. 2 is another schematic flow chart of a method for predicting the commodity sales of an offline store according to an embodiment of the present disclosure;
fig. 3 is a schematic structural diagram of an offline store commodity sales amount prediction apparatus according to an embodiment of the present application.
Detailed Description
In order to make the technical solutions of the present application better understood, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
For easy understanding, referring to fig. 1, an embodiment of a method for predicting sales of off-line stores provided by the present application includes:
The first historical sales data of the commodities at least comprise historical sales volume, holidays, weather, information of offline stores and the like, wherein the quantity of the commodities is multiple, and the commodities comprise target commodities.
And 102, performing target feature extraction and hierarchical classification on the first historical sales data to obtain a plurality of training sets.
And performing target feature extraction on the first historical sales data, wherein the target features at least comprise seasonal features, holiday features, weather features and store features (trade circles, periodicity), and also comprise promotion plan features and the like, and the target features can be stored by adopting a wide table. The method comprises the steps of marking the types of first historical sales data for distinguishing different types of data, hierarchically classifying the first historical sales data from commodity attributes such as regions, commodity classes, sales volumes, volatility, sales frequency and the like, marking the first historical sales data in a hierarchical marking mode, marking each dimension of the same first historical sales data from large to small in range, marking the first historical sales data in a hierarchical marking mode, wherein the first historical sales data can be marked in a region label such as province, city and store, marking the commodity class labels in a large, medium and small mode, hierarchically classifying the first historical sales data according to the labels to obtain a plurality of training sets, each training set is in a hierarchical category, each training set comprises a plurality of training samples, each training sample comprises target characteristics and sales volumes, and the characteristics of each training set are more prominent by hierarchically classifying the data, and the training sets have more consistent characteristics such as randomness, randomness and the like of the training sets, Seasonality and tendency, so that the model is selected more pertinently, and the prediction accuracy of the prediction model can be improved.
And 103, inputting each training set into a plurality of models to be trained respectively for training to obtain a plurality of prediction models.
Considering that different types of data are suitable for different models, and a single model cannot be used for matching the optimal model for each type of data, in the embodiment of the application, each training set is respectively input into a plurality of models to be trained for training, when the models to be trained reach a convergence condition, a plurality of prediction models are obtained, the number of the prediction models is the product of the number of the training sets and the number of the models to be trained, wherein the models to be trained at least comprise machine learning models and time sequence models, and the machine learning models comprise support vector machines, random forests, convolutional neural networks and other models.
And 104, performing target feature extraction on the acquired second historical sales data of the target commodity to obtain first features, and respectively inputting the first features into each prediction model to obtain a plurality of predicted sales volumes.
The time period of the second historical sales data from the current time is less than the time period of the first historical sales data from the current time, and the second historical sales data at least comprises data such as historical sales volume, holidays, weather and information of offline stores. And extracting target characteristics of the second historical sales data to obtain first characteristics, and respectively inputting the first characteristics into each prediction model to obtain a plurality of predicted sales.
And 105, calculating an error value between the actual sales volume of the target commodity and each predicted sales volume, and taking the prediction model corresponding to the minimum error value as the optimal prediction model.
And calculating an error value between the actual sale of the target commodity and each predicted sale amount to evaluate each prediction model, wherein the error value can be a weighted average absolute percentage error value, a mean square error value or a mean absolute error value, and the prediction model with the minimum error value is selected as the optimal prediction model of the target commodity. The best prediction model is selected by retesting the prediction model through the target commodity, the best prediction model can be selected by a data retest method objectively and scientifically, and the risk of judgment by artificial experience is reduced.
And 106, acquiring third history sales data of the target commodity, extracting target features of the third history sales data to obtain second features, inputting the second features into the optimal prediction model, and outputting the sales volume of the target commodity in a preset time period.
When the third historical sales data is sales data of about 1 month, the third historical sales data can be sales data of about 1 month, holidays of about one month, weather of about one month, offline store information and the like. And after third history sales data of the target commodity are obtained, target feature extraction is carried out on the third history sales data to obtain second features, the second features are input into the optimal prediction model obtained through the back measurement, and the sales volume of the target commodity in a preset time period is output.
According to the method for predicting the commodity sales volume of the off-line stores, first historical sales data of commodities of a plurality of off-line stores are obtained, target feature extraction and hierarchical classification are carried out on the first historical sales data to obtain a plurality of training sets, and through the hierarchical classification, the features of each training set are outstanding, each training set has more consistent features, and a model to be trained can be better trained; each data set is respectively input into a plurality of models to be trained for training, and a plurality of models to be trained such as a machine learning model and a time sequence model are used for training, so that the problem of low prediction accuracy of a single model is avoided; the method comprises the steps of extracting target characteristics of the obtained second historical sales data of the target commodity, inputting the second historical sales data into each prediction model to obtain a plurality of predicted sales, measuring the error values of the predicted sales and the actual sales back and forth by calculating, selecting the optimal prediction model suitable for the target commodity, predicting the sales of the target commodity in a preset time period by using the optimal prediction model, and measuring back by using the second historical sales data of the target commodity, so that the prediction model most suitable for the target commodity is selected, the prediction accuracy is improved, and the technical problem that the prediction accuracy is low due to the fact that a single prediction model is adopted in the existing sales prediction method is solved.
For ease of understanding, referring to fig. 2, the present application further provides another embodiment of a method for predicting sales of store products offline, including:
The first historical sales data of the commodity at least comprises historical sales volume, holidays, weather, offline store information and the like. The preprocessing comprises data screening, data missing value filling or abnormal value modification, and the data are preprocessed, so that the quality of the first historical sales data can be improved, and the prediction performance of the prediction model is improved.
And performing target feature extraction on the first historical sales data, wherein the target features at least comprise seasonal features, holiday features, weather features and store features (trade circles, periodicity), and also comprise promotion plan features and the like, and the target features can be stored by adopting a wide table. The method comprises the steps of marking the types of first historical sales data to distinguish different types of data, classifying the first historical sales data from regions, commodity classes, sales volumes, volatility and sales frequency, marking the first historical sales data in a layered marking mode, marking labels on all dimensions of the same first historical sales data from large to small in range, wherein the labels on the regions can be provinces, cities and stores, the commodity classes can be large, medium and small, classifying the first historical sales data in a layered mode according to the labels to obtain a plurality of training sets, each training set is a layered class, each training set comprises a plurality of training samples, each training sample comprises target features and sales volumes, and the data are classified in a layered mode to enable the features of the training sets to be more prominent and have more consistent features inside, such as randomness, seasonality and the like of the training sets, The trend enables the model to be selected more specifically, and the prediction accuracy of the prediction model can be improved.
And 203, inputting each training set into a plurality of models to be trained respectively for training to obtain a plurality of prediction models.
Considering that different types of data are suitable for different models, and a single model cannot be used for matching the optimal model for each type of data, in the embodiment of the application, each training set is respectively input into a plurality of models to be trained for training, when the models to be trained reach a convergence condition, a plurality of prediction models are obtained, the number of the prediction models is the product of the number of the training sets and the number of the models to be trained, wherein the models to be trained at least comprise machine learning models and time sequence models, and the machine learning models comprise support vector machines, random forests, convolutional neural networks and other models.
And step 204, acquiring second historical sales data of the target commodity, and preprocessing the second historical sales data.
The time period of the second historical sales data from the current time is less than the time period of the first historical sales data from the current time, and the second historical sales data at least comprises data such as historical sales volume, holidays, weather and information of offline stores. Preprocessing includes data screening, data missing value population, or outlier modification.
And 205, performing target feature extraction on the preprocessed second historical sales data to obtain first features, and inputting the first features into each prediction model respectively to obtain a plurality of predicted sales volumes.
And performing target feature extraction on the preprocessed second historical sales data to obtain first features, and inputting the first features into each prediction model respectively to obtain a plurality of predicted sales volumes.
And step 206, calculating an error value between the actual sales volume of the target commodity and each predicted sales volume, and taking the prediction model corresponding to the minimum error value as the optimal prediction model.
And calculating an error value between the actual sale of the target commodity and each predicted sale amount to evaluate each prediction model, wherein the error value can be a weighted average absolute percentage error value, a mean square error value or a mean absolute error value, and the prediction model with the minimum error value is selected as the optimal prediction model of the target commodity. The best prediction model is selected by retesting the prediction model through the target commodity, the best prediction model can be selected by a data retest method objectively and scientifically, and the risk of judgment by artificial experience is reduced.
And step 207, acquiring third history sales data of the target commodity, and preprocessing the third history sales data.
When the third historical sales data is sales data of about 1 month, the third historical sales data can be sales data of about 1 month, holidays of about one month, weather of about one month, offline store information and the like. And after the third history sales data of the target commodity is acquired, preprocessing the third history sales data, wherein the preprocessing comprises data screening, data missing value filling or abnormal value modification.
And 208, extracting target characteristics of the preprocessed third history sales data to obtain second characteristics, inputting the second characteristics into the optimal prediction model, and outputting the sales volume of the target commodity in a preset time period.
And extracting target characteristics from the preprocessed third history sales data to obtain second characteristics, inputting the second characteristics into the optimal prediction model obtained by the back measurement, and outputting the sales volume of the target commodity in a preset time period.
For easy understanding, referring to fig. 3, the present application provides an offline store commodity sales predicting apparatus, including:
the first acquiring unit 301 is configured to acquire first historical sales data of commodities of a plurality of offline stores.
The classification unit 302 is configured to perform target feature extraction and hierarchical classification on the first historical sales data to obtain a plurality of training sets.
The training unit 303 is configured to input each training set to a plurality of models to be trained, respectively, to perform training, so as to obtain a plurality of prediction models, where the number of the prediction models is a product of the number of the training sets and the number of the models to be trained, and the models to be trained at least include a machine learning model and a time series model.
And the extracting unit 304 is configured to perform target feature extraction on the acquired second historical sales data of the target product to obtain first features, and input the first features into each prediction model respectively to obtain a plurality of predicted sales volumes, where a time period of the second historical sales data from the current time is less than a time period of the first historical sales data from the current time.
The calculation unit 305 is configured to calculate an error value between the actual sales volume of the target product and each predicted sales volume, and use a prediction model corresponding to the smallest error value as the optimal prediction model.
The prediction unit 306 is configured to obtain third historical sales data of the target commodity, perform target feature extraction on the third historical sales data to obtain a second feature, input the second feature into the optimal prediction model, and output a sales volume of the target commodity in a preset time period, where a time period from the third historical sales data to the current time is less than a time period from the second historical sales data to the current time.
As a further improvement, the method further comprises the following steps:
and a preprocessing unit 307 for preprocessing the first historical sales data, wherein the preprocessing includes data filtering, data missing value filling or abnormal value modification.
As a further refinement, the target feature comprises at least a seasonal feature, a holiday feature, a weather feature, or a store feature.
As a further refinement, the error value comprises a weighted average absolute percentage error value, a mean square error value, or a mean absolute error value.
The application also provides an offline store commodity sales predicting device, which comprises a processor and a memory;
the memory is used for storing the program codes and transmitting the program codes to the processor;
the processor is used for executing the offline store commodity sales amount prediction method in the offline store commodity sales amount prediction method embodiment according to instructions in the program code.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, 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 application 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 above embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should 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 such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions in the embodiments of the present application.
Claims (10)
1. A method for predicting commodity sales of off-line stores is characterized by comprising the following steps:
acquiring first historical sales data of commodities of a plurality of offline stores;
performing target feature extraction and hierarchical classification on the first historical sales data to obtain a plurality of training sets;
respectively inputting each training set into a plurality of models to be trained for training to obtain a plurality of prediction models, wherein the number of the prediction models is the product of the number of the training sets and the number of the models to be trained, and the models to be trained at least comprise a machine learning model and a time sequence model;
performing target feature extraction on the acquired second historical sales data of the target commodity to obtain first features, and respectively inputting the first features into each prediction model to obtain a plurality of predicted sales volumes, wherein the time period of the second historical sales data from the current time is less than the time period of the first historical sales data from the current time;
calculating an error value of the actual sales volume of the target commodity and each predicted sales volume, and taking the prediction model corresponding to the minimum error value as an optimal prediction model;
and acquiring third historical sales data of the target commodity, extracting target features of the third historical sales data to obtain second features, inputting the second features into the optimal prediction model, and outputting the sales volume of the target commodity in a preset time period, wherein the time period of the third historical sales data from the current time is less than the time period of the second historical sales data from the current time.
2. The method for predicting the commodity sales of the off-line store according to claim 1, wherein the target feature extraction of the first historical sales data further comprises:
preprocessing the first historical sales data, wherein the preprocessing comprises data screening, data missing value filling or abnormal value modification.
3. The method of predicting the quantity of commodity sales for off-line stores according to claim 1, wherein the target features include at least seasonal features, holiday features, weather features, or store features.
4. The method for predicting the commodity sales of the off-line store according to claim 1, wherein the hierarchically classifying the first historical sales data includes:
and hierarchically classifying the first historical sales data based on commodity attributes, wherein the commodity attributes at least comprise regions, commodity categories, sales volumes, volatility and sales frequency.
5. The method of predicting the sales of commodities from off-line stores according to claim 1, wherein the error value comprises a weighted mean absolute percentage error value, a mean square error value, or a mean absolute error value.
6. An offline store commodity sales amount prediction device, comprising:
the system comprises a first acquisition unit, a second acquisition unit and a display unit, wherein the first acquisition unit is used for acquiring first historical sales data of commodities of a plurality of offline stores;
the classification unit is used for carrying out target feature extraction and hierarchical classification on the first historical sales data to obtain a plurality of training sets;
the training unit is used for inputting each training set into a plurality of models to be trained respectively for training to obtain a plurality of prediction models, the number of the prediction models is the product of the number of the training sets and the number of the models to be trained, and the models to be trained at least comprise a machine learning model and a time sequence model;
the extraction unit is used for extracting target characteristics of the obtained second historical sales data of the target commodity to obtain first characteristics, the first characteristics are respectively input into each prediction model to obtain a plurality of predicted sales volumes, and the time period of the second historical sales data from the current time is less than the time period of the first historical sales data from the current time;
a calculation unit, configured to calculate an error value between an actual sales volume of the target product and each predicted sales volume, and use the prediction model corresponding to the smallest error value as an optimal prediction model;
and the prediction unit is used for acquiring third historical sales data of the target commodity, extracting target characteristics of the third historical sales data to obtain second characteristics, inputting the second characteristics into the optimal prediction model, and outputting the sales volume of the target commodity in a preset time period, wherein the time period of the third historical sales data from the current time is less than the time period of the second historical sales data from the current time.
7. The offline store commodity sales prediction apparatus according to claim 6, further comprising:
and the preprocessing unit is used for preprocessing the first historical sales data, and the preprocessing comprises data screening, data missing value filling or abnormal value modification.
8. The offline store commodity sales prediction apparatus according to claim 6, wherein the target characteristics include at least a seasonal characteristic, a holiday characteristic, a weather characteristic, or a store characteristic.
9. The off-line store commodity sales prediction device of claim 6, wherein the error value comprises a weighted mean absolute percentage error value, a mean square error value, or a mean absolute error value.
10. An offline store commodity sales prediction apparatus, comprising a processor and a memory;
the memory is used for storing program codes and transmitting the program codes to the processor;
the processor is configured to execute the offline store commodity sales prediction method of any one of claims 1-5 according to instructions in the program code.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010614259.0A CN111724211A (en) | 2020-06-30 | 2020-06-30 | Offline store commodity sales prediction method, device and equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010614259.0A CN111724211A (en) | 2020-06-30 | 2020-06-30 | Offline store commodity sales prediction method, device and equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111724211A true CN111724211A (en) | 2020-09-29 |
Family
ID=72570296
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010614259.0A Pending CN111724211A (en) | 2020-06-30 | 2020-06-30 | Offline store commodity sales prediction method, device and equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111724211A (en) |
Cited By (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112241904A (en) * | 2020-10-23 | 2021-01-19 | 浙江集享电子商务有限公司 | Commodity sales prediction method, commodity sales prediction device, computer equipment and storage medium |
CN112308624A (en) * | 2020-11-09 | 2021-02-02 | 深圳市苍泓科技有限公司 | Internet-based online commodity sales prediction method and equipment |
CN112501755A (en) * | 2020-12-28 | 2021-03-16 | 浙江肯特科技股份有限公司 | Incoming line clamping machine control management system for electric blanket washing marks |
CN112581182A (en) * | 2020-12-24 | 2021-03-30 | 山东新北洋信息技术股份有限公司 | Sales management method and system for automatic vending equipment |
CN112669070A (en) * | 2020-12-29 | 2021-04-16 | 上海小零网络科技有限公司 | Method for predicting store sales volume to develop stores |
CN112669069A (en) * | 2020-12-29 | 2021-04-16 | 上海小零网络科技有限公司 | Sales prediction method for off-line retail store |
CN112686705A (en) * | 2020-12-31 | 2021-04-20 | 上海画龙信息科技有限公司 | Method and device for predicting sales effect data and electronic equipment |
CN112785057A (en) * | 2021-01-21 | 2021-05-11 | 上海东普信息科技有限公司 | Component prediction method, device, equipment and storage medium based on exponential smoothing |
CN112884520A (en) * | 2021-03-09 | 2021-06-01 | 拉扎斯网络科技(上海)有限公司 | Object processing method, device, electronic equipment, storage medium and program product |
CN113408626A (en) * | 2021-06-22 | 2021-09-17 | 浙江省烟草公司宁波市公司 | Merchant classification method |
CN113434611A (en) * | 2021-07-08 | 2021-09-24 | 宁波权智科技有限公司 | Sales classification method and device based on artificial intelligence |
CN113435541A (en) * | 2021-07-22 | 2021-09-24 | 创优数字科技(广东)有限公司 | Method and device for planning product classes, storage medium and computer equipment |
CN113487440A (en) * | 2021-07-26 | 2021-10-08 | 联仁健康医疗大数据科技股份有限公司 | Model generation method, health insurance claim settlement determination method, device, equipment and medium |
CN113554455A (en) * | 2021-06-30 | 2021-10-26 | 杭州拼便宜网络科技有限公司 | Store commodity analysis method and device based on artificial intelligence and storage medium |
CN113610575A (en) * | 2021-08-06 | 2021-11-05 | 浙江吉利控股集团有限公司 | Product sales prediction method and prediction system |
CN113723989A (en) * | 2021-08-10 | 2021-11-30 | 杉数科技(北京)有限公司 | Product sales prediction method and related equipment |
CN113743643A (en) * | 2021-02-05 | 2021-12-03 | 北京京东振世信息技术有限公司 | Method, device, equipment and medium for determining commodity data prediction accuracy |
CN113763055A (en) * | 2021-10-25 | 2021-12-07 | 金蝶软件(中国)有限公司 | Prediction method for store commodity distribution and related equipment |
CN113807758A (en) * | 2020-12-02 | 2021-12-17 | 北京京东振世信息技术有限公司 | Data generation method and device |
CN113888235A (en) * | 2021-10-22 | 2022-01-04 | 创优数字科技(广东)有限公司 | Training method of sales forecasting model, sales forecasting method and related device |
WO2022271794A1 (en) * | 2021-06-25 | 2022-12-29 | Z2 Cool Comics Llc | Semi-autonomous advertising systems and methods |
CN115736162A (en) * | 2022-11-21 | 2023-03-07 | 河南农业大学 | Method and system for preparing prefabricated dishes based on intelligent data analysis |
CN115841343A (en) * | 2022-12-16 | 2023-03-24 | 广州飞狮数字科技有限公司 | Method and device for determining sales amount |
CN115994788A (en) * | 2023-03-20 | 2023-04-21 | 北京永辉科技有限公司 | Data processing analysis method and device |
CN117436936A (en) * | 2023-12-19 | 2024-01-23 | 湖南兴盛优选电子商务有限公司 | Sales prediction and BOM finished product processing system and method |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102968670A (en) * | 2012-10-23 | 2013-03-13 | 北京京东世纪贸易有限公司 | Method and device for predicting data |
CN108734330A (en) * | 2017-04-24 | 2018-11-02 | 北京京东尚科信息技术有限公司 | Data processing method and device |
US20180341898A1 (en) * | 2017-05-24 | 2018-11-29 | Sears Brands, L.L.C. | Demand forecast |
CN109214601A (en) * | 2018-10-31 | 2019-01-15 | 四川长虹电器股份有限公司 | Household electric appliances big data Method for Sales Forecast method |
CN109493151A (en) * | 2019-01-10 | 2019-03-19 | 哈步数据科技(上海)有限公司 | Method for Sales Forecast method and system |
CN109509030A (en) * | 2018-11-15 | 2019-03-22 | 北京旷视科技有限公司 | Method for Sales Forecast method and its training method of model, device and electronic system |
CN109559163A (en) * | 2018-11-16 | 2019-04-02 | 广州麦优网络科技有限公司 | A kind of model building method and sales forecasting method based on machine learning |
CN109741082A (en) * | 2018-11-29 | 2019-05-10 | 杭州览众数据科技有限公司 | A kind of seasonal merchandise needing forecasting method based on Time Series |
CN110390425A (en) * | 2019-06-20 | 2019-10-29 | 阿里巴巴集团控股有限公司 | Prediction technique and device |
CN111008858A (en) * | 2019-10-24 | 2020-04-14 | 清华大学 | Commodity sales prediction method and system |
CN111028016A (en) * | 2019-12-12 | 2020-04-17 | 腾讯科技(深圳)有限公司 | Sales data prediction method and device and related equipment |
-
2020
- 2020-06-30 CN CN202010614259.0A patent/CN111724211A/en active Pending
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102968670A (en) * | 2012-10-23 | 2013-03-13 | 北京京东世纪贸易有限公司 | Method and device for predicting data |
CN108734330A (en) * | 2017-04-24 | 2018-11-02 | 北京京东尚科信息技术有限公司 | Data processing method and device |
US20180341898A1 (en) * | 2017-05-24 | 2018-11-29 | Sears Brands, L.L.C. | Demand forecast |
CN109214601A (en) * | 2018-10-31 | 2019-01-15 | 四川长虹电器股份有限公司 | Household electric appliances big data Method for Sales Forecast method |
CN109509030A (en) * | 2018-11-15 | 2019-03-22 | 北京旷视科技有限公司 | Method for Sales Forecast method and its training method of model, device and electronic system |
CN109559163A (en) * | 2018-11-16 | 2019-04-02 | 广州麦优网络科技有限公司 | A kind of model building method and sales forecasting method based on machine learning |
CN109741082A (en) * | 2018-11-29 | 2019-05-10 | 杭州览众数据科技有限公司 | A kind of seasonal merchandise needing forecasting method based on Time Series |
CN109493151A (en) * | 2019-01-10 | 2019-03-19 | 哈步数据科技(上海)有限公司 | Method for Sales Forecast method and system |
CN110390425A (en) * | 2019-06-20 | 2019-10-29 | 阿里巴巴集团控股有限公司 | Prediction technique and device |
CN111008858A (en) * | 2019-10-24 | 2020-04-14 | 清华大学 | Commodity sales prediction method and system |
CN111028016A (en) * | 2019-12-12 | 2020-04-17 | 腾讯科技(深圳)有限公司 | Sales data prediction method and device and related equipment |
Non-Patent Citations (2)
Title |
---|
刘竹林等: "《统计学学习指导》", 31 October 2014, 中国科学技术大学出版社, pages: 199 * |
马慧民: "《智能新零售——数据智能时代的零售业变革》", 31 January 2019, 中国铁道出版社, pages: 118 - 119 * |
Cited By (36)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112241904A (en) * | 2020-10-23 | 2021-01-19 | 浙江集享电子商务有限公司 | Commodity sales prediction method, commodity sales prediction device, computer equipment and storage medium |
CN112308624A (en) * | 2020-11-09 | 2021-02-02 | 深圳市苍泓科技有限公司 | Internet-based online commodity sales prediction method and equipment |
CN113807758A (en) * | 2020-12-02 | 2021-12-17 | 北京京东振世信息技术有限公司 | Data generation method and device |
CN113807758B (en) * | 2020-12-02 | 2023-11-03 | 北京京东振世信息技术有限公司 | Data generation method and device |
CN112581182A (en) * | 2020-12-24 | 2021-03-30 | 山东新北洋信息技术股份有限公司 | Sales management method and system for automatic vending equipment |
CN112581182B (en) * | 2020-12-24 | 2023-04-07 | 山东新北洋信息技术股份有限公司 | Sales management method and system for automatic vending equipment |
CN112501755A (en) * | 2020-12-28 | 2021-03-16 | 浙江肯特科技股份有限公司 | Incoming line clamping machine control management system for electric blanket washing marks |
CN112501755B (en) * | 2020-12-28 | 2022-04-08 | 浙江肯特科技股份有限公司 | Incoming line clamping machine control management system for electric blanket washing marks |
CN112669070A (en) * | 2020-12-29 | 2021-04-16 | 上海小零网络科技有限公司 | Method for predicting store sales volume to develop stores |
CN112669069A (en) * | 2020-12-29 | 2021-04-16 | 上海小零网络科技有限公司 | Sales prediction method for off-line retail store |
CN112686705B (en) * | 2020-12-31 | 2024-02-27 | 上海画龙信息科技有限公司 | Method and device for predicting sales effect data and electronic equipment |
CN112686705A (en) * | 2020-12-31 | 2021-04-20 | 上海画龙信息科技有限公司 | Method and device for predicting sales effect data and electronic equipment |
CN112785057A (en) * | 2021-01-21 | 2021-05-11 | 上海东普信息科技有限公司 | Component prediction method, device, equipment and storage medium based on exponential smoothing |
CN112785057B (en) * | 2021-01-21 | 2023-09-01 | 上海东普信息科技有限公司 | Method, device, equipment and storage medium for predicting quantity of parts based on exponential smoothing |
CN113743643B (en) * | 2021-02-05 | 2023-11-03 | 北京京东振世信息技术有限公司 | Method, device, equipment and medium for determining commodity data prediction accuracy |
CN113743643A (en) * | 2021-02-05 | 2021-12-03 | 北京京东振世信息技术有限公司 | Method, device, equipment and medium for determining commodity data prediction accuracy |
CN112884520A (en) * | 2021-03-09 | 2021-06-01 | 拉扎斯网络科技(上海)有限公司 | Object processing method, device, electronic equipment, storage medium and program product |
CN113408626A (en) * | 2021-06-22 | 2021-09-17 | 浙江省烟草公司宁波市公司 | Merchant classification method |
US20220414706A1 (en) * | 2021-06-25 | 2022-12-29 | Z2 Cool Comics Llc | Semi-Autonomous Advertising Systems and Methods |
WO2022271794A1 (en) * | 2021-06-25 | 2022-12-29 | Z2 Cool Comics Llc | Semi-autonomous advertising systems and methods |
CN113554455A (en) * | 2021-06-30 | 2021-10-26 | 杭州拼便宜网络科技有限公司 | Store commodity analysis method and device based on artificial intelligence and storage medium |
CN113434611A (en) * | 2021-07-08 | 2021-09-24 | 宁波权智科技有限公司 | Sales classification method and device based on artificial intelligence |
CN113435541A (en) * | 2021-07-22 | 2021-09-24 | 创优数字科技(广东)有限公司 | Method and device for planning product classes, storage medium and computer equipment |
CN113487440A (en) * | 2021-07-26 | 2021-10-08 | 联仁健康医疗大数据科技股份有限公司 | Model generation method, health insurance claim settlement determination method, device, equipment and medium |
CN113610575A (en) * | 2021-08-06 | 2021-11-05 | 浙江吉利控股集团有限公司 | Product sales prediction method and prediction system |
CN113610575B (en) * | 2021-08-06 | 2024-03-08 | 浙江吉利控股集团有限公司 | Product sales prediction method and prediction system |
CN113723989A (en) * | 2021-08-10 | 2021-11-30 | 杉数科技(北京)有限公司 | Product sales prediction method and related equipment |
CN113888235A (en) * | 2021-10-22 | 2022-01-04 | 创优数字科技(广东)有限公司 | Training method of sales forecasting model, sales forecasting method and related device |
CN113763055B (en) * | 2021-10-25 | 2024-03-19 | 金蝶软件(中国)有限公司 | Store commodity distribution prediction method and related equipment |
CN113763055A (en) * | 2021-10-25 | 2021-12-07 | 金蝶软件(中国)有限公司 | Prediction method for store commodity distribution and related equipment |
CN115736162A (en) * | 2022-11-21 | 2023-03-07 | 河南农业大学 | Method and system for preparing prefabricated dishes based on intelligent data analysis |
CN115841343B (en) * | 2022-12-16 | 2024-01-30 | 广州飞狮数字科技有限公司 | Sales limit determining method and device |
CN115841343A (en) * | 2022-12-16 | 2023-03-24 | 广州飞狮数字科技有限公司 | Method and device for determining sales amount |
CN115994788B (en) * | 2023-03-20 | 2023-06-30 | 北京永辉科技有限公司 | Data processing analysis method and device |
CN115994788A (en) * | 2023-03-20 | 2023-04-21 | 北京永辉科技有限公司 | Data processing analysis method and device |
CN117436936A (en) * | 2023-12-19 | 2024-01-23 | 湖南兴盛优选电子商务有限公司 | Sales prediction and BOM finished product processing system and method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111724211A (en) | Offline store commodity sales prediction method, device and equipment | |
CN106933956B (en) | Data mining method and device | |
US20200074486A1 (en) | Information processing system, information processing device, prediction model extraction method, and prediction model extraction program | |
CN109284948A (en) | Logistics object selection method, logistics object selection device and electronic device | |
US20190034945A1 (en) | Information processing system, information processing method, and information processing program | |
CN113469730A (en) | Customer repurchase prediction method and device based on RF-LightGBM fusion model under non-contract scene | |
CN110008977B (en) | Clustering model construction method and device | |
CN110335090A (en) | Replenishing method and system, electronic equipment based on Sales Volume of Commodity forecast of distribution | |
CN110321914A (en) | A kind of Oil Quality Analysis managing and control system | |
CN113704389A (en) | Data evaluation method and device, computer equipment and storage medium | |
CN114493641A (en) | Information display method and device, electronic equipment and computer readable medium | |
CN116610821A (en) | Knowledge graph-based enterprise risk analysis method, system and storage medium | |
CN115759885A (en) | Material sampling inspection method and device based on distributed material supply | |
EP3493082A1 (en) | A method of exploring databases of time-stamped data in order to discover dependencies between the data and predict future trends | |
CN111984637B (en) | Missing value processing method and device in data modeling, equipment and storage medium | |
CN114862188A (en) | Analysis system for agricultural product e-commerce data | |
CN115392992A (en) | Commodity recommendation method, terminal device and computer-readable storage medium | |
CN111353794A (en) | Data processing method, supply chain scheduling method and device | |
CN117150389B (en) | Model training method, carrier card activation prediction method and equipment thereof | |
CN113095870B (en) | Prediction method, prediction device, computer equipment and storage medium | |
CN116485020B (en) | Supply chain risk identification early warning method, system and medium based on big data | |
CN110688273A (en) | Classification model monitoring method and device, terminal and computer storage medium | |
CN111754056B (en) | Model evaluation method and related device | |
CN116739836B (en) | Restaurant data analysis method and system based on knowledge graph | |
CN115269409A (en) | Method and device for evaluating operation quality of terminal, electronic equipment and storage medium |
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 | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20201120 Address after: Room 011, first floor, no.2429, Xingang East Road, Haizhu District, Guangzhou City, Guangdong Province (office only) Applicant after: CHUANGYOU digital technology (Guangdong) Co.,Ltd. Address before: Room 205, 2f office, No.156, nanshanzui Road, Hengqin, Zhuhai City, Guangdong Province Applicant before: MINISO (HENGQIN) ENTERPRISE MANAGEMENT Co.,Ltd. |
|
TA01 | Transfer of patent application right | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200929 |
|
RJ01 | Rejection of invention patent application after publication |