KR102294125B1 - Method and system for predicting product demand using artificial intelligence - Google Patents

Abstract

A method of operating a server for providing a product demand prediction service includes the following steps of: receiving characteristic information of a franchise store from a franchise store management device; generating a list of products to be provided to the franchise store based on the characteristic information of the franchise store; predicting a demand value of each of the products; receiving feedback about the list of the products to be provided to the franchise store and the predicted demand value of each of the products from the franchise store management device; and updating the list of the products to be provided to the franchise store and the predicted demand value of each of the products by reflecting the feedback. The characteristic information of the franchise store can include at least one piece of information among information about consumers who visit the franchise store, and demand and stock information about the products managed in the franchise store and each of the products managed in the franchise store. Therefore, the present invention is capable of providing a service for predicting products optimal for franchise stores.

Description

Product demand forecasting method and system applying artificial intelligence {METHOD AND SYSTEM FOR PREDICTING PRODUCT DEMAND USING ARTIFICIAL INTELLIGENCE}

The present invention is a product demand prediction method and system to which artificial intelligence is applied, and more specifically, a product demand prediction method and system capable of providing demand prediction information for products optimized to each affiliated store based on information provided from affiliated stores is about

Franchising means that the head office grants the franchisee the right to provide goods or services using the trademark, trade name, product, or business operation method of the franchise head office within a certain region for a certain period of time, and the franchisee grants the goods or the above services. It means a business method that pays royalties corresponding to an initial subscription fee or a certain percentage of sales in return for obtaining the right to provide

That is, the head office may have a structure to manage a plurality of affiliated stores, and in order to maximize the profits of the plurality of affiliated stores, products optimized for each of the affiliated stores should be provided, and it is preferable that each of the affiliated stores maintain a stock of products. However, when each of the affiliated stores excessively stocks products, excessively stocked products may result in a waste of management resources.

Accordingly, the head office managing a plurality of affiliated stores may require a system and service that provides a demand value of each of the products optimized for each of the affiliated stores.

Patent Publication 10-2016-0142789, 2016.12.13

An object of the present invention to solve the above problems is to derive a correlation between products based on characteristic information of products obtained from a plurality of affiliated stores, and to add or change products to affiliated stores based on the derived relationship. It is to provide a method and system for providing a service that can predict optimized products to affiliated stores by recommending it to the manager of

Another object of the present invention to solve the above problems is to derive the correlation between products and the trend of change in the demand value of the product, etc. based on the characteristic information of the products obtained from a plurality of affiliated stores, and the derived relationship and to provide a method and system for providing a service capable of predicting the demand for each of the products to be provided to the affiliated store based on trend information and the like.

In order to achieve the above object, a method of operating a server for providing a product demand prediction service according to a first embodiment of the present invention includes: receiving characteristic information of an affiliated store from an affiliated store management device; generating a list of products to be provided to the affiliated store based on the characteristic information of the affiliated store; predicting a demand value of each of the products; receiving, from the affiliate store management device, feedback on a list of products to be provided to the affiliated store and a predicted demand value of each of the products; and updating a list of products to be provided to the affiliated store and a demand value of each of the products by reflecting the feedback, wherein the characteristic information of the affiliated store includes information about consumers who have visited the affiliated store, information about the customer who has visited the affiliated store, and managing the affiliated store. It may include at least one piece of information among products and demand information and inventory information of each of the products managed by the affiliated store.

Here, the generating of the list of products to be provided to the affiliated store may include: calculating a suitability of each of the products managed by the affiliated store; calculating information on products having a degree of suitability greater than or equal to a preset value; and determining a new product to be additionally provided in the affiliated store based on information on products having a suitability greater than or equal to the preset value.

에 의해 산출될 수 있다. Here, in the step of calculating the suitability of each of the products managed by the affiliated store, the products managed by the affiliated store by reflecting the demand information, inventory information, and environmental information of each of the products managed by the affiliated store Each fit can be calculated, and the fit is

can be calculated by

Here, the suitability of each of the products managed by the affiliated store may be determined based on trend information of demand information and inventory information of each of the products managed by the affiliated store.

Here, the operating method of the system includes: collecting characteristic information of a plurality of affiliated stores received from a plurality of affiliated store management devices; classifying the collected characteristic information according to products managed by the affiliated store; and analyzing a correlation between products managed by the affiliated store based on the collected characteristic information, wherein the new product has a positive correlation with products having a suitability greater than or equal to the preset value. It can be a product.

Here, the operating method of the system includes: selecting products having a degree of suitability less than a preset value from among the products managed by the affiliated store; and determining a product capable of replacing products having a degree of fitness less than the preset value, wherein the product capable of replacing products having a fitness of less than the preset value is less than the preset value. It may be a product with a good fit and a product with a negative correlation.

에 의해 산출될 수 있다.Here, predicting the demand value of each of the products may include calculating the demand value of the product at the affiliate store at time #n by applying a prediction model including a plurality of preset techniques; calculating an error between the demand values calculated by applying the plurality of techniques of the prediction model and the measured demand value of the product at time #n; determining one of the plurality of techniques based on the calculated error value; And by applying the determined one technique, the demand value of the product at time #n+1 can be predicted, and the demand value of the product at the time #n+1 is

can be calculated by

Here, the plurality of techniques of the prediction model include at least one of a prediction technique using a moving average method, a prediction technique using an exponential smoothing method, a prediction technique using an autocorrelation model, a time series analysis technique, and a prediction technique using an artificial neural network. techniques may be included.

Here, the operating method of the system may further include updating a weight vector of the artificial neural network based on the calculated error value.

Here, the step of updating the list of products to be provided to the affiliated store and the demand value of each of the products by reflecting the feedback includes each of the predicted products based on the characteristic information of the affiliated store included in the feedback. calculating an error of the demand value; The method may further include updating the prediction model based on the error of the demand value of each of the products.

According to the present invention, a product demand prediction method and system to which artificial intelligence is applied derives a relationship between products based on characteristic information of products obtained from a plurality of affiliated stores, and adds or changes the relationship based on the derived relationship. By recommending products to the manager of the affiliated store, it is possible to provide a service for predicting optimized products to the affiliated store.

According to the present invention, the product demand prediction method and system to which artificial intelligence is applied derives the correlation between products and the change trend of the demand value of the product, etc. based on the characteristic information of the products obtained from a plurality of affiliated stores. It is possible to provide a service capable of predicting the demand for each product to be provided to the affiliated store based on relationship and trend information.

1 is a block diagram illustrating an embodiment of the structure of a system for providing a product demand forecasting service.
2 is a flowchart illustrating one embodiment of an operation of a system for providing a product demand forecasting service.
3 is a flow diagram illustrating one embodiment of operation of a system for generating a list of products to be offered to a merchant.
4 is a conceptual diagram illustrating an embodiment of an operation result of a system for deriving a correlation between products to be provided to an affiliated store.
5 is a flowchart illustrating an embodiment of an operation of a system for predicting a demand value of each product to be provided to an affiliated store.
6 is a flowchart illustrating one embodiment of operation of a system for updating a list of products to be provided to a merchant and a demand value based on feedback from the merchant.
7 is an exemplary diagram illustrating a hardware configuration of a device included in the system for providing the product demand prediction service according to FIG. 1 .
8 is a conceptual diagram illustrating an embodiment of an artificial neural network included in a product demand prediction server.

Since the present invention can have various changes and can have various embodiments, specific embodiments are illustrated in the drawings and described in detail. However, this is not intended to limit the present invention to specific embodiments, and it should be understood to include all modifications, equivalents and substitutes included in the spirit and scope of the present invention.

Terms such as first, second, etc. may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, a second component may also be referred to as a first component. and/or includes a combination of a plurality of related listed items or any of a plurality of related listed items.

When a component is referred to as being “connected” or “connected” to another component, it is understood that the other component may be directly connected or connected to the other component, but other components may exist in between. it should be On the other hand, when it is said that a certain element is "directly connected" or "directly connected" to another element, it should be understood that no other element is present in the middle.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise. In the present application, terms such as “comprise” or “have” are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in a commonly used dictionary should be interpreted as having a meaning consistent with the meaning in the context of the related art, and should not be interpreted in an ideal or excessively formal meaning unless explicitly defined in the present application. does not

Hereinafter, preferred embodiments of the present invention will be described in more detail with reference to the accompanying drawings. In describing the present invention, in order to facilitate the overall understanding, the same reference numerals are used for the same components in the drawings, and duplicate descriptions of the same components are omitted.

1 is a block diagram illustrating an embodiment of the structure of a system for providing a product demand forecasting service.

Referring to FIG. 1 , a system for providing a product demand forecasting service may include a product demand forecasting server 110 and a merchant management device 130 connected to the product demand forecasting server 110 through a network 120 . have. The system for providing the product demand forecasting service may be a system providing the optimized product demand forecasting service to the affiliated store through a wired or wireless communication network.

The product demand prediction server 110 may store information necessary to operate a product demand prediction service. The information necessary to operate the product demand forecasting service may include products and information on each of the products that the operator of the product demand forecasting service may provide. And the product demand prediction server 110 may store the existing characteristic information of each of the products in the affiliate store.

The network 120 may mean a connection structure for exchanging information between the product demand prediction server 110 and the affiliate store management device 130 . Network 120 is Internet (Internet), LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), 3G, 4G, LTE (long term evolution) , VoLTE (voice over LTE), 5G NR (new radio) Wi-Fi (wireless-fidelity), Bluetooth (Bluetooth), NFC, RFID (radio frequency identification) home network, IoT (internet of things), etc. may include.

The affiliate store management device 130 may be connected to the product demand prediction server 110 through the network 120 . In addition, the affiliated store management device 130 may include a user device such as a computer, a tablet, or a smart phone. The affiliate store management device 130 may receive information necessary to provide a product demand forecasting service from the product demand forecasting server 110 . Information necessary to provide a product demand prediction service may include information such as available products, ingredient information of each product, main customer base, efficacy, and purchase history of existing consumers.

The affiliate store management device 130 may display the received information for providing the product demand prediction service through an output device. In addition, the affiliated store management device 130 may receive information on the product demand forecasting service from the product demand forecasting server 110 and display a product demand forecasting result more optimized for the affiliated store.

The service provided through the present invention may be a service for managing the demand for each product at the affiliate store, but it will be apparent that the service provided by the present invention is not limited thereto. The operation of the server of the system for providing the product demand forecasting service may be as described below.

2 is a flowchart illustrating one embodiment of an operation of a system for providing a product demand forecasting service. Referring to FIG. 2 , the product demand prediction server of the system for providing the product demand prediction service may provide the optimized product demand prediction service to each of the affiliated stores based on the characteristic information of the affiliated store provided from the affiliated store.

The product demand prediction server may secure information on products from an externally connected server. For example, the externally connected server may include a server of a product manufacturer and/or a server such as an online market selling products.

That is, the product demand prediction server may provide a product demand prediction service optimized for each of the affiliated stores by using information on products obtained in advance based on the characteristic information of the affiliated stores. A specific operation of the product demand prediction server for providing the optimized product demand prediction service to each of the affiliated stores may be as described below.

In step S210, the product demand prediction server may receive the characteristic information of the affiliate store. The affiliate store management device may transmit characteristic information of the affiliate store to the product demand forecast server through a network connected to the product demand forecast server. For example, the characteristic information of the affiliated store may include information about consumers who have visited the affiliated store, products managed by the affiliated store, and demand information and inventory information of products managed by the affiliated store. That is, the product demand prediction server may receive characteristic information of the affiliated store from the affiliated store management device.

In step S220, the product demand prediction server may generate a list of products to be provided to the affiliate store. A specific operation of the product demand prediction server generating a list of products to be provided to the affiliated store based on the characteristic information of the affiliated store obtained from the affiliated store management device may be described below.

3 is a flow diagram illustrating one embodiment of operation of a system for generating a list of products to be offered to a merchant.

Referring to FIG. 3 , the product demand prediction server of the system for providing a product demand prediction service may generate a list of products optimized for each of the affiliated stores based on the characteristic information of each of the affiliated stores provided from a plurality of affiliated stores. have.

In step S221, the product demand prediction server may calculate the suitability of each of the products managed by the affiliate store. The product demand prediction server may calculate the suitability of each of the products managed by the affiliated store by reflecting demand information, inventory information, and environmental information of each of the products managed by the affiliated store. The suitability of each product managed by the affiliate store can be calculated by [Equation 1], [Equation 1] is as follows.

[Equation 1]

는

제품의 적합도를 지시하고,

는

시점에서의

제품의 수요 값을 지시하고,

는

시점에서의

제품의 재고 값을 지시하고,

은 표본을 획득한 기간을 지시할 수 있다. 그리고

는

시점에 따른 가중치 값을 지시할 수 있으며,

는

제품에 대한 가맹점의 환경 정보를 지시할 수 있다. 여기서,

는 최근 시점에 가까울수록 큰 값일 수 있으며,

및

는 미리 설정된 값의 가중치일 수 있다. In Equation 1,

Is

indicate the suitability of the product;

Is

at the point

indicate the demand value of the product,

Is

at the point

indicate the stock value of the product;

may indicate the period during which the sample was obtained. and

Is

It is possible to indicate the weight value according to the time point,

Is

It is possible to instruct the environmental information of the affiliated store for the product. here,

may have a larger value the closer to the latest time,

and

may be a weight of a preset value.

In step S222 , the product demand prediction server may calculate information on products having a suitability greater than or equal to a preset value among products managed by the affiliate store. The product demand prediction server may classify a product having a suitability greater than or equal to a preset value as a product to be provided in the affiliate store when the fitness of the product is greater than or equal to a preset value based on the fitness of each of the products managed by the affiliate store.

In step S223 , the product demand prediction server may determine a new product to be additionally provided in the affiliated store based on information on products having a suitability greater than or equal to a preset value. The product demand prediction server may determine a new product to be additionally provided in the affiliate store based on the correlation between the products.

In step S224 , when the fitness level of the product is less than a preset value, the product demand prediction server may classify a product having a fitness degree of less than or equal to the preset value as a product to be excluded from the affiliated store.

In step S225 , the product demand prediction server may determine a product that can replace products having a degree of fitness less than a preset value based on information on products having a degree of fitness less than a preset value. The product demand prediction server may determine a product that can replace the existing product based on the correlation between the products.

4 is a graph for deriving a correlation between the suitability of each of the products provided to the affiliated store and the products.

Referring to Figure 4, the product demand prediction server of the system for providing the product demand forecasting service, in order to derive the degree of suitability of each of the products provided to the affiliated store and the correlation between the products, demand information of each of the products managed by the affiliated store, At least one piece of inventory information may be used. In this embodiment, an embodiment based on demand information of each product will be described, but it will be apparent that the embodiment is not limited to the above embodiment.

The product demand prediction server may derive a demand trend for each product in the affiliate store based on information (eg, demand information) of each product obtained from the affiliate store, and based on the direction of the trend line of the derived demand With this, it is possible to calculate the suitability of each product managed by the affiliated store. For example, among products A to E, the demand values of products A, B, and C tend to increase overall, and the demand values for product D overall decrease. Accordingly, the product demand prediction server may determine that the product A, product B, and product C have high suitability for affiliated stores, and may determine that product D has low suitability for affiliate stores.

In addition, the product demand prediction server may derive a correlation between products managed by the affiliated store based on the direction of the derived demand trend line. For example, since product A and product D have opposite increasing/decreasing trend values of demand values for each section, the product demand prediction server may derive the correlation between product A and product D as a negative correlation. In addition, since product C and product E have similar increase/decrease trend values for each section, the product demand prediction server may derive the correlation between product C and product E as a positive correlation. Therefore, when there is another affiliated store that has only C in the past, the product demand prediction server may determine the new product to additionally provide E at the other affiliated store. And it can be determined that product D is a substitute product for product A.

Again, referring to FIG. 2 , in step S230, the product demand prediction server may predict a quantity value of each of the products included in the determined list. A specific operation in which the product demand prediction server predicts the demand value of each product may be as described below.

5 is a flowchart illustrating an embodiment of an operation of a system for predicting a demand value of each product to be provided to an affiliated store.

Referring to FIG. 5 , the product demand prediction server of the system for providing a product demand prediction service may predict the demand value of products optimized for each of the affiliated stores based on the characteristic information of each of the affiliated stores provided from a plurality of affiliated stores. have.

In step S231, the product demand prediction server may calculate the demand value of the product at time #n by applying a plurality of techniques. The product demand prediction server may apply a prediction model including a plurality of preset techniques to calculate a product demand prediction value at the affiliate store at time #n. Here, the plurality of techniques of the prediction model may include at least one of a prediction technique using a moving average method, a prediction technique using an exponential smoothing method, a prediction technique using an autocorrelation model, a time series analysis technique, and a prediction technique using an artificial neural network. techniques may be included.

In step S232, the product demand prediction server may calculate the error of the calculated values by applying a plurality of techniques. The product demand prediction server may calculate an error between the predicted product demand values at the affiliated store at the time #n calculated by applying a plurality of techniques of the prediction model and the actual demand value of the product at the affiliated store at the #n time point.

In step S233, the product demand prediction server may determine one technique to be used in the prediction model based on the error of the demand value. The product demand prediction server predicts the demand with the lowest error with the actual demand value of the product at the merchant at the time #n among the demand forecast values of the product at the affiliated store at the #n time point calculated by applying a plurality of techniques of the prediction model. The technique that calculated the value can be determined as one technique to be used in the predictive model.

[Equation 2]

은 수요 예측 모델에 포함된 수요 예측 기법들의 개수를 지시할 수 있고,

은

가지 수요 예측 기법들 중 결정된 하나의 기법일 수 있다. 그리고

는

시점에서의 제품의 수요 값을 지시하며,

는

번째 수요 예측 기법을 적용하여 예측된

시점에서의 제품의 수요 값일 수 있다. 즉, 제품 수요 예측 서버는

가지 수요 예측 기법을 적용하여 예측된

시점에서의 제품의 수요 값 중 실제 제품의 수요 값과의 차이가 가장 적은

번째 예측 기법을 이용하여,

시점에서의 제품의 수요 값(

)을 예측할 수 있다. in Equation 2

may indicate the number of demand forecasting techniques included in the demand forecasting model,

silver

It may be one method determined among the various demand forecasting methods. and

Is

It indicates the demand value of the product at the time,

Is

The second demand forecasting technique is applied

It may be a demand value of the product at a point in time. In other words, the product demand forecasting server

Forecasted by applying various demand forecasting techniques

The difference between the actual product demand value and the actual product demand value at the point in time is the smallest.

Using the second prediction method,

The demand value of the product at the time (

) can be predicted.

In step S234 , the product demand prediction server may predict the demand value of the product at the time #n+1 by applying the determined one technique.

In step S235, the product demand prediction server may update the weight vector of the artificial neural network of the prediction model based on the error of the demand value. The product demand prediction server may update the weight vector of the artificial neural network by backpropagating the error of the demand value.

An administrator who manages a plurality of affiliated stores may provide products to each of the affiliated stores based on information on products included in the list generated by the product demand prediction server and a demand value of each of the products.

Referring again to FIG. 2 , in step S240 , the product demand prediction server may receive a list of products and a feedback for each demand value of the products from the affiliate store management device. The feedback information for the list of products and the demand value of each of the products is obtained after a certain time has elapsed from the time of obtaining the products provided based on the list of products and the demand value of each of the products generated by the product demand prediction server. It may include characteristic information such as demand information and inventory information of each of the products of the affiliated store.

In step S250 , the product demand prediction server may update a list of products and a demand value of each of the products based on the feedback. The product demand prediction server may receive feedback information on a list of products of the affiliated store and a demand value of each of the products from the affiliated store management device, and may update characteristic information of the affiliated store based on the received feedback information. That is, the product demand prediction server may update information about the consumer who has visited the affiliated store, the products managed by the affiliated store, and demand information and inventory information of each of the products managed by the affiliated store. In addition, the product demand prediction server may update a list of products to be provided to the affiliated store and a demand value of each of the products based on the updated characteristic information of the affiliated store. A specific operation of the product demand prediction server for updating the list of products to be provided to the affiliated store and the demand value based on the feedback from the affiliated store may be as described below.

6 is a flowchart illustrating one embodiment of operation of a system for updating a list of products to be provided to a merchant and a demand value based on feedback from the merchant.

Referring to FIG. 6 , the product demand prediction server of the system for providing the product demand prediction service may update the list of products to be provided to the affiliated store and the demand value based on the feedback information provided from the affiliated store.

In step S251, the product demand prediction server may update the list of products to be provided to the affiliated store based on the feedback received from the affiliated store management device. The product demand prediction server can update the demand trend of each product at the affiliated store based on the existing information and feedback of each of the products obtained from the affiliated stores, and manage it at the affiliated store based on the direction of the trend line of the derived demand Each of the products can be updated.

In step S252 , the product demand prediction server may calculate an error of the previously predicted product demand value. The product demand prediction server may apply [Equation 2] to calculate the error of the demand value of the product previously predicted based on the data at time n+1.

In step S253 , the product demand prediction server may update one technique to be used in the prediction model based on the error of the demand value. When the error between the previously predicted product demand value and the actual product demand value included in the feedback information exceeds a preset range, the product demand prediction server uses Analysis techniques can be changed.

In step S254, the product demand prediction server may update the weight vector of the artificial neural network of the prediction model based on the error of the demand value. The product demand prediction server may update the weight vector of the artificial neural network by backpropagating the error of the demand value.

Referring back to FIG. 2 , in step S260 , the product demand prediction server may provide products to the affiliated store based on the updated list of products and the demand value.

The product demand prediction server includes an operation of predicting a list of products to be provided to the affiliated store and a demand value of each of the products, an operation of acquiring feedback information from the affiliated store device, and a list of products to be provided to the affiliated store and each of the products based on the feedback information The operations of updating the demand value of may be repeated. Accordingly, the product demand prediction server may continuously update the product recommendation model optimized for the affiliated store and the demand prediction model of the products to increase the accuracy of product recommendation for the affiliated store.

7 is an exemplary diagram illustrating a hardware configuration of a device included in the system for providing the product demand prediction service according to FIG. 1 .

Referring to FIG. 7 , a device 700 included in a system for providing a product demand forecasting service includes at least one processor 710 and at least one processor 710 a command to instruct to perform at least one step. It may include a memory 720 for storing instructions.

Here, the at least one processor 710 may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed. can Each of the memory 720 and the storage device 760 may be configured as at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory 720 may be configured as at least one of a read only memory (ROM) and a random access memory (RAM).

In addition, the device 700 included in the system for providing the product demand prediction service may include a transceiver 730 that performs communication through a wireless network. In addition, the device 700 included in the system for providing the product demand prediction service may further include an input interface device 740 , an output interface device 750 , a storage device 760 , and the like. Each component included in the device 700 included in the system for providing the product demand prediction service may be connected by a bus 770 to communicate with each other.

8 is a conceptual diagram illustrating an embodiment of an artificial neural network included in a product demand prediction server.

Referring to FIG. 8 , the artificial neural network of the product demand prediction server may include a plurality of layers, and each layer may include a plurality of artificial nodes. In addition, the artificial neural network of the product demand prediction server may include a plurality of adaptive weight vectors, and artificial nodes included in respective layers of the artificial neural network may be connected by the weight vectors. The artificial neural network of the product demand prediction server may include a plurality of layers and artificial nodes according to preset parameters.

The artificial neural network of the product demand prediction server may include four layers, and specifically, an input layer 810 , an output layer 840 , and hidden layers 820 and 830 . ) may be included. The input layer 810 may be a layer into which pre-processed characteristic information of the affiliated store is input.

In addition, the output layer 840 may be a layer that outputs the predicted product demand value based on the characteristic information of the affiliated store input to the input layer. In addition, two of the four layers 820 and 830 may be hidden layers 820 and 830 . The hidden layers 820 and 830 may be connected to the input layer 810 through a weight vector, or the hidden layers 820 and 830 may be connected to the output layer 840 through a weight vector.

The product demand prediction server may perform a learning operation of updating the weight vectors of the artificial neural network. The learning operation may be performed by a multi-layer perceptron classifier included in the product demand prediction server.

The product demand forecasting server may further include a multi-layer perceptron classifier. The multilayer perceptron classifier can train an artificial neural network through a preset learning algorithm. The learning algorithm may include a learning algorithm such as a supervised learning algorithm and a non-supervised learning algorithm.

The methods according to the present invention may be implemented in the form of program instructions that can be executed by various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the computer-readable medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the art of computer software.

Examples of computer-readable media include hardware devices specially configured to store and carry out program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as at least one software module to perform the operations of the present invention, and vice versa.

Although it has been described with reference to the above embodiments, it will be understood by those skilled in the art that various modifications and changes can be made to the present invention without departing from the spirit and scope of the present invention as set forth in the following claims. will be able

700: device
710: processor
720: memory
730: transceiver
740: input interface device
750: output interface device
760: storage device

Claims (10)

In the operating method of a server for providing a product demand forecasting service,
receiving characteristic information of an affiliated store from an affiliated store management device;
generating a list of products to be provided to the affiliated store based on the characteristic information of the affiliated store;
predicting a demand value of each of the products;
receiving, from the affiliate store management device, feedback on a list of products to be provided to the affiliated store and a predicted demand value of each of the products; and
and updating a list of products to be provided to the affiliated store and a demand value of each of the products by reflecting the feedback,
The characteristic information of the affiliated store is,
It includes at least one of information about the consumer who has visited the affiliated store, the products managed by the affiliated store, and demand information and inventory information of each of the products managed by the affiliated store,
The step of generating a list of products to be provided to the affiliated store includes:
calculating the suitability of each of the products managed by the affiliated store;
calculating information on products having a degree of suitability greater than or equal to a preset value; and
determining a new product to be additionally provided in the affiliated store based on information on products having a degree of suitability greater than or equal to the preset value;
The step of calculating the suitability of each of the products managed by the affiliate store is,
By reflecting the demand information, inventory information, and environmental information of each of the products managed by the affiliated store, the suitability of each of the products managed by the affiliated store is calculated, and the suitability is calculated by Equation 1 below become,
The suitability of each of the products managed by the affiliated store is,
It is determined based on trend information of demand information and inventory information of each of the products managed by the affiliated store,
The step of generating a list of products to be provided to the affiliated store includes:
collecting characteristic information of a plurality of affiliated stores received from a plurality of affiliated store management devices;
classifying the collected characteristic information according to products managed by the affiliated store; and
and analyzing a correlation between products managed by the affiliated store based on the collected characteristic information,
The new product is
It is a product having a positive correlation with products having a degree of fitness greater than or equal to the preset value,
The step of generating a list of products to be provided to the affiliated store includes:
selecting products having a degree of suitability less than a preset value from among the products managed by the affiliate store; and
Further comprising the step of determining a product that can replace products having a degree of suitability less than the preset value,
A product that can replace products with a degree of conformity less than the preset value,
a product having a negative correlation with a product having a degree of fitness less than the preset value;
The step of predicting the demand value of each of the products,
calculating a demand value of a product at a merchant at time #n by applying a prediction model including a plurality of preset techniques;
calculating an error between the demand values calculated by applying the plurality of techniques of the prediction model and the measured demand value of the product at time #n;
determining one of the plurality of techniques based on the calculated error value; and
Applying the determined one technique, including the step of predicting the demand value of the product at the time #n+1, the demand value of the product at the time #n+1 is calculated by Equation 2 below, how it works.
(Equation 1)

(here,

Is

is the degree of conformity of the product,

Is

at the point

is the demand value of the product,

Is

at the point

is the stock value of the product,

is the period during which the sample was obtained,

Is

It is the weight value according to the time point,

Is

It is the environmental information of the affiliated store for the product)
(Equation 2)

(here,

is the number of demand forecasting techniques included in the demand forecasting model,

silver

It is one method determined among the demand forecasting methods of branches,

Is

is the demand value of the product at the point in time,

Is

The second demand forecasting technique is applied

is the demand value of the product at the point in time,

Is

It is the demand value of the product at the point in time) delete delete delete delete delete delete The method according to claim 1,
The plurality of techniques of the predictive model are
An operation method comprising at least one of a prediction method using a moving average method, a prediction method using an exponential smoothing method, a prediction method using an autocorrelation model, a time series analysis method, and a prediction method using an artificial neural network. 9. The method of claim 8,
The method further comprising the step of updating a weight vector of the artificial neural network based on the calculated error value. 9. The method of claim 8,
The step of updating the list of products to be provided to the affiliated store and the demand value of each of the products by reflecting the feedback,
calculating an error of a demand value of each of the products predicted based on the characteristic information of the affiliated store included in the feedback;
The method further comprising the step of updating the predictive model based on the error of the demand value of each of the products.

Images