KR102491688B1 - Control method of electronic apparatus for determining predictive modeling of direction of financial investment products - Google Patents

Abstract

Disclosed is an electronic device control method. The control method includes the steps of: extracting a plurality of first combinations including a first number of parameters randomly selected from among a plurality of parameters; identifying the performance of each of a plurality of first combinations; selecting at least one first combination having excellent performance from among the plurality of first combinations; extracting a plurality of second combinations in which the selected first combination is dimensionally reduced; and identifying the performance of each of a plurality of second combinations. Therefore, it is possible to increase the accuracy of prediction.

Description

Electronic device control method for determining direction prediction modeling method of financial investment product { CONTROL METHOD OF ELECTRONIC APPARATUS FOR DETERMINING PREDICTIVE MODELING OF DIRECTION OF FINANCIAL INVESTMENT PRODUCTS

The present disclosure relates to a control method of an electronic device that models a predictive method of a parameter such as a direction of a financial investment product, and more particularly, to an electronic device that searches for an optimal combination that can be used for prediction within a plurality of parameters. It is about the control method of.

In predicting various parameters such as stock prices, various modeling methods for feature selection have been used.

However, in that accuracy, roc-auc, f1, etc., as well as performance evaluation scales related to hit rates, CAGR (compound annual return), Sharpe Ratio (Sharp index), and MDD (maximum drop) should be additionally considered, existing modeling Time-series predictive modeling of finance-related parameters was not smooth using the method alone.

Publication No. 10-2021-0143460 (feature recommendation device and its feature recommendation method)

The present disclosure provides a control method of an electronic device capable of increasing prediction accuracy by searching for an optimal combination to be applied to an artificial intelligence algorithm among a plurality of parameters.

The objects of the present disclosure are not limited to the above-mentioned objects, and other objects and advantages of the present disclosure not mentioned above can be understood by the following description and will be more clearly understood by the embodiments of the present disclosure. Further, it will be readily apparent that the objects and advantages of the present disclosure may be realized by means of the instrumentalities and combinations indicated in the claims.

A control method of an electronic device according to an embodiment of the present disclosure includes extracting a plurality of first combinations including a first number of parameters randomly selected from among a plurality of parameters, and parameters included in each of the plurality of first combinations. Identifying the performance of each of the plurality of first combinations by inputting values of these into an artificial intelligence algorithm, selecting at least one first combination having excellent performance among the plurality of first combinations based on the identified performance extracting a plurality of second combinations in which the selected first combination is dimensionally reduced, inputting values of parameters included in each of the plurality of second combinations into an artificial intelligence algorithm, and each of the plurality of second combinations It includes identifying the performance of

In the extracting of a plurality of first combinations, a preset number of first combinations including the first number of parameters randomly selected from among the plurality of parameters may be obtained. The predetermined number may be smaller than the number of all cases in which the first number of parameters may be selected from the plurality of parameters.

Also, the preset number may be set according to the number of the plurality of parameters.

In the step of extracting a plurality of second combinations, a plurality of second combinations matching the number of all cases in which the second number of parameters among the parameters included in the selected first combination may be selected may be obtained. . The second number may be smaller than the first number.

In the step of identifying the performance of each of the plurality of first combinations, an artificial intelligence algorithm consisting of at least one of random forest, XGBoost (Extreme Gradient Boosting), KNN (K-Nearest Neighbor), and SVM (Support Vector Machine) is used A performance of each of the plurality of first combinations may be identified. In addition, the step of identifying the performance of each of the plurality of second combinations may include an artificial intelligence algorithm composed of at least one of random forest, extreme gradient boosting (XGBoost), K-nearest neighbor (KNN), and support vector machine (SVM). It is possible to identify the performance of each of the plurality of second combinations by utilizing.

Meanwhile, the control method of the electronic device may include selecting at least one second combination having excellent performance among the plurality of second combinations as a target combination, inputting values of parameters constituting the target combination into an artificial intelligence algorithm, and , predicting the value of at least one parameter.

In this case, the control method of the electronic device may include extracting a plurality of first combinations including a first number of parameters randomly selected from among the plurality of parameters when the accuracy of the predicted parameter value is less than a threshold value; The method may further include selecting at least one target combination based on the re-extracted plurality of first combinations, and comparing performances of the selected target combination and the additionally selected target combination.

In addition, the control method of the electronic device may include a first predicting step of predicting a value of at least one parameter based on values of parameters constituting the target combination of first data, and constituting the target combination of second data. The second predicting step of predicting the value of at least one parameter based on the values of the parameters, when the difference between the accuracy of the first predicting step and the accuracy of the second predicting step is a predetermined value or more, among the plurality of second combinations It may also include changing the target combination with at least one other second combination.

The method for controlling an electronic device according to the present disclosure may select parameters contributing to improved prediction accuracy by automatically extracting a parameter combination having relatively excellent performance among numerous combinations that may be extracted through a plurality of parameters.

1 is a block diagram for explaining the configuration and functions of an electronic device according to an embodiment of the present disclosure;
2 is a flowchart for explaining a control method of an electronic device according to an embodiment of the present disclosure;
3 is a diagram for explaining an operation of primarily selecting a combination with excellent performance by an electronic device according to an embodiment of the present disclosure;
4 is a diagram for explaining an operation of secondarily selecting a combination with excellent performance by an electronic device according to an embodiment of the present disclosure; and
5 is a block diagram for explaining the configuration of an electronic device according to various embodiments of the present disclosure.

Prior to a detailed description of the present disclosure, the method of describing the present specification and drawings will be described.

First, terms used in the present specification and claims are general terms in consideration of functions in various embodiments of the present disclosure. However, these terms may vary depending on the intention of a technician working in the art, legal or technical interpretation, and the emergence of new technologies. In addition, some terms are arbitrarily selected by the applicant. These terms may be interpreted as the meanings defined in this specification, and if there is no specific term definition, they may be interpreted based on the overall content of this specification and common technical knowledge in the art.

In addition, the same reference numerals or numerals in each drawing attached to this specification indicate parts or components that perform substantially the same function. For convenience of description and understanding, the same reference numerals or symbols are used in different embodiments. That is, even if all components having the same reference numerals are shown in a plurality of drawings, the plurality of drawings do not mean one embodiment.

Also, in the present specification and claims, terms including ordinal numbers such as “first” and “second” may be used to distinguish between elements. These ordinal numbers are used to distinguish the same or similar components from each other, and the meaning of the term should not be construed as being limited due to the use of these ordinal numbers. For example, the order of use or arrangement of elements associated with such ordinal numbers should not be limited by the number. If necessary, each ordinal number may be used interchangeably.

In this specification, singular expressions include plural expressions unless the context clearly dictates otherwise. In this application, the terms "comprise" or "consist of" are intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, but one or more other It should be understood that the presence or addition of features, numbers, steps, operations, components, parts, or combinations thereof is not precluded.

In the embodiments of the present disclosure, terms such as “module,” “unit,” and “part” are terms used to refer to components that perform at least one function or operation, and these components are hardware or software. It may be implemented or implemented as a combination of hardware and software. In addition, a plurality of "modules", "units", "parts", etc. are integrated into at least one module or chip, except for cases where each of them needs to be implemented with separate specific hardware, so that at least one processor can be implemented as

Also, in an embodiment of the present disclosure, when a part is said to be connected to another part, this includes not only a direct connection but also an indirect connection through another medium. In addition, the meaning that a certain part includes a certain component means that it may further include other components without excluding other components unless otherwise stated.

1 is a block diagram for explaining the configuration and functions of an electronic device according to an embodiment of the present disclosure.

Referring to FIG. 1 , an electronic device 100 may include a memory 110 and a processor 120 . The electronic device 100 may correspond to various electronic devices including one or more computers.

For example, the electronic device 100 may be a server of a financial institution, a securities company, or other securities-related service provider. In this case, the electronic device 100 may communicate with various user terminals (eg, smart phone, tablet PC, notebook PC, desktop PC) through at least one application or web page.

As another example, the electronic device 100 may correspond to a terminal device such as a smart phone, a tablet PC, a notebook PC, or a desktop PC.

The memory 110 is an operating system (OS) for controlling the overall operation of the components of the electronic device 100 and a component for storing at least one instruction or data related to the components of the electronic device 100. .

The memory 110 may include non-volatile memory such as ROM and flash memory, and may include volatile memory such as DRAM. Also, the memory 110 may include a hard disk, a solid state drive (SSD), and the like.

The memory 110 may include information on various parameters to be monitored or predicted.

For example, the memory 110 may include information on parameters related to various financial investment products such as finance and securities. As a specific example, the parameters may include various indicators related to the price and trading of securities, such as stock price, direction of change in stock price, magnitude of change in stock price, number of securities, trading volume, trading price, individual trading volume, institutional trading volume, and foreigner trading volume. can In addition, the parameters may include operating profit, assets, capital, and the like of securities-related companies. However, the type of parameter is not limited to the above-described example, and various other parameters may be utilized as a matter of course.

The processor 120 is a component for controlling the electronic device 100 as a whole. Specifically, the processor 130 may perform operations according to various embodiments of the present disclosure by executing at least one instruction stored in the memory 110 while being connected to the memory 110 .

The processor 120 may include a general-purpose processor such as a CPU, an AP, or a digital signal processor (DSP), a graphics-only processor such as a GPU or a vision processing unit (VPU), or an artificial intelligence-only processor such as an NPU. An artificial intelligence-only processor may be designed as a hardware structure specialized for training or use of a specific artificial intelligence model.

The processor 120 may monitor values of a plurality of parameters in real time and/or periodically. To this end, the processor 120 may communicate with various external devices such as a user terminal, a financial institution server, and a securities company server through a communication unit.

The processor 120 may perform preprocessing on values of a plurality of parameters related to various economic indicators, and may perform normalization on data. For example, the processor 120 may remove missing values from among the values of a plurality of parameters and perform normalization through a rate-of-change function, a log function, an exponential function, or the like for unit unification of each parameter. For example, the processor 120 may obtain a normalized parameter value (S _T ) by performing preprocessing according to the following equation.

Here, R _T represents the change rate of the parameter (: economic indicator data) with respect to time (T). R _T may be set according to Equation 2 below.

Also, the processor 120 may remove overlapping parameters among a plurality of parameters and retain only one parameter among the parameters.

Also, the processor 120 may define and set at least one new parameter by utilizing at least one parameter among a plurality of parameters.

For example, as shown in the following equation, the processor 120 may generate a parameter corresponding to the average annual return (GR) based on parameters such as an end value, a beginning value, and a time frame (N) related to an investment period.

In addition, as shown in the following equation, the processor 120 calculates the portfolio's profit (R _p ), the portfolio's risk-related index (R _f = risk-free rate), the portfolio's deviation from the excess profit (σ _p ), etc. A parameter (SR) corresponding to a separate indicator may be created.

In addition, as shown in the following equation, the processor 120 may generate a parameter M corresponding to a separate indicator based on the peak value of the observation period and the trough value of the observation period.

The processor 120 may generate various combinations including at least one parameter among a plurality of parameters and identify performance of the generated combinations. Specifically, the processor 120 may evaluate performance based on accuracy of prediction as a result of inputting values of parameters in the combination to an artificial intelligence algorithm. In this case, the accuracy of the prediction may be measured according to the return on investment based on the values of the parameters included in the corresponding combination, but is not limited thereto.

However, when it is assumed that the number of a plurality of parameters is n, the load is excessive to identify performance for combinations corresponding to the number of all cases (2 ⁿ -1 combinations). In this regard, the processor 120 may randomly select only a certain number (eg, k) of combinations to perform performance evaluation, and specific embodiments will be described later through the drawings below.

Referring to FIG. 1 , the processor 120 may control a first combination extraction module 121 , a performance evaluation module 122 , and a second combination extraction module 123 . These modules may be implemented through software and/or hardware.

The first combination extraction module 121 is a module for extracting one or more first combinations in which some of a plurality of parameters are combined. The first combination extraction module 121 may extract a first combination by randomly selecting a first number of parameters from among a plurality of parameters, and as this process is repeated several times, a plurality of first combinations may be obtained.

The performance evaluation module 122 is a module for evaluating the performance of combinations (a first combination and a second combination) extracted from a plurality of parameters.

The performance evaluation module 122 identifies the performance of each combination by utilizing an artificial intelligence algorithm consisting of at least one of random forest, extreme gradient boosting (XGBoost), K-Nearest Neighbor (KNN), and support vector machine (SVM). can

For example, the performance evaluation module 122 may determine performance depending on how consistently the values of the parameters constituting the combination are correlated with the stock price (the value of the parameter to be predicted), but is not limited thereto. .

The second combination extraction module 123 is a module for extracting a plurality of second combinations based on at least one of the plurality of first combinations extracted by the first combination extraction module 121 .

Specifically, the second combination extraction module 123 may extract a plurality of second combinations obtained by dimensionally reducing the first combination. For example, when the first combination has 10 parameters, a plurality of second combinations including only 4 of them may be extracted.

An operation of the electronic device 100 that searches for an optimal combination will be described in more detail through the following drawings.

2 is a flowchart illustrating a control method of an electronic device according to an embodiment of the present disclosure.

Referring to FIG. 2 , the electronic device 100 may extract a plurality of first combinations including a first number of parameters randomly selected from among a plurality of parameters (S210).

Specifically, the first combination extraction module 121 may obtain a predetermined number of first combinations including a first number (eg, 5, 10, 15, etc.) of parameters randomly selected from among a plurality of parameters. there is.

Here, the preset number may be smaller than the number of all cases in which the first number of parameters may be selected from a plurality of parameters. Since it may take excessive load and operation time of the electronic device 100 to search for all searchable cases, as described above, only some randomly extracted first combinations are extracted and their performances are compared.

The preset number may be set according to the number of a plurality of parameters. For example, the preset number may be set to be larger as the number of the plurality of parameters increases, but is not limited thereto.

FIG. 3 is a diagram for explaining an operation of extracting a plurality of first combinations from which 10 (first number) parameters among a plurality of parameters can be selected.

Referring to FIG. 3 , among all possible combinations of 10 parameters among a plurality of parameters, only some combinations (first combinations) may be randomly extracted and actually searched.

Specifically, the performance evaluation module 122 of the electronic device 100 may identify the performance of each of the extracted first combinations (S220).

The performance evaluation module 122 utilizes an artificial intelligence algorithm consisting of at least one of random forest, extreme gradient boosting (XGBoost), K-Nearest Neighbor (KNN), and support vector machine (SVM) to perform performance of each of a plurality of first combinations. can identify.

Then, the electronic device 100 may select at least one first combination having excellent performance from among the (extracted) plurality of first combinations (S230).

Specifically, referring to FIG. 3 , the electronic device 100 may arrange a plurality of first combinations in descending order of performance, and may select at least one first combination having a relatively excellent performance. For example, the electronic device 100 may select five first combinations having the best performance among a plurality of extracted first combinations.

Then, the electronic device 100 may extract a plurality of second combinations obtained by dimensionally reducing each of the selected first combinations (S240).

Specifically, the second combination extraction module 123 may obtain a plurality of second combinations matching the number of all cases in which the second number of parameters among the parameters included in the selected first combination may be selected. Here, the second number may be smaller than the aforementioned first number.

Referring to FIG. 4, in a first combination composed of a first number (: 10) of parameters, pluralities matching the number of all cases in which only four (: second number) of parameters can be selected. A second combination of can be obtained (dimension reduction).

Then, the electronic device 100 may identify the performance of the extracted plurality of second combinations (S250).

For example, the performance evaluation module 122 may identify the performance of each of the plurality of second combinations by inputting values of parameters included in each of the plurality of second combinations to the artificial intelligence algorithm. In this case, the artificial intelligence algorithm may be composed of at least one of random forest, extreme gradient boosting (XGBoost), K-nearest neighbor (KNN), and support vector machine (SVM), but is not limited thereto. Meanwhile, although the performance of the first combination and the performance of the second combination are both identified through one performance evaluation module 122 through the above-described embodiment, since the dimensions of the first combination and the second combination are different, each Of course, an independent performance evaluation module may be utilized.

As a result, referring to FIG. 4 , the second optimal combination (parameters A, D, H, and U) having the best performance may be selected.

In this case, the electronic device 100 may set at least one second combination having the best performance as the target combination.

The target combination refers to a combination of parameters actually used in an operation of the electronic device 100 that predicts a value of at least one parameter.

Specifically, the electronic device 100 inputs the values of the parameters constituting the target combination into an artificial intelligence algorithm, and inputs the value of at least one parameter (ex. stock price, direction of change in stock price, amount of change in stock price) Predictable. This operation may correspond to the operation of the prediction module 124 to be described later in FIG. 5 .

Meanwhile, the electronic device 100 may monitor the accuracy of prediction performed through the target combination. Also, the electronic device 100 may change the target combination according to the prediction accuracy.

Specifically, when the accuracy of the parameter value predicted according to the target combination is less than the threshold value, the electronic device 100 may perform the above-described process of FIG. 2 again. In this case, since a plurality of first combinations different from the previously performed process of FIG. 2 may be extracted, a new target combination may be additionally selected.

If the existing target combination and the additionally selected target combination are the same, the electronic device 100 may maintain the existing target combination.

On the other hand, when the existing target combination and the additionally selected target combination are different, the electronic device 100 may compare performances of the existing target combination and the additionally selected target combination. As a result, target combinations with better performance can be utilized for future predictions.

Meanwhile, the electronic device 100 may monitor whether the target combination has disproportionately high accuracy only for specific data.

Specifically, the electronic device 100 may collect first data and second data that are distinguished from each other. Here, the first data and the second data correspond to values of the plurality of parameters extracted from different databases.

In this case, the electronic device 100 may perform a first prediction step of predicting a value of at least one parameter based on values of parameters constituting the target combination among the first data. Here, the electronic device 100 may perform prediction by inputting values of parameters constituting the target combination into an artificial intelligence algorithm.

In addition, the electronic device 100 may also perform a second predicting step of predicting the value of at least one parameter based on the values of the parameters constituting the target combination among the second data. Similarly, artificial intelligence algorithms can be utilized.

Here, the electronic device 100 may compare the accuracy of the first prediction step and the accuracy of the second prediction step. If the difference between accuracies is greater than or equal to a predetermined value, the electronic device 100 may change the target combination to at least one other second combination.

Meanwhile, when a plurality of first combinations are extracted according to the process S210, the processor 120 of the electronic device 100 may monitor the real-time workload of the processor 120.

In this case, when the real-time workload of the processor 120 is greater than the threshold workload, the electronic device 100 may classify the plurality of first combinations into two or more groups.

Specifically, when a plurality of first combinations including the first number of parameters randomly selected from among the plurality of parameters are extracted (S210), the electronic device 100 divides the plurality of first combinations into two or more groups (first group, second group, etc.).

In this case, the electronic device 100 may acquire and provide a first target combination by performing steps S220 to S250 with respect to the first combinations of the first group, and may use the first target combination to obtain a value of at least one parameter. predictions can be made.

Also, after the first target combination is obtained, the electronic device 100 may acquire the second target combination by performing processes S220 to S250 on the first combinations of the second group.

Here, the electronic device 100 compares the performance of the first target combination and the second target combination, selects one target combination having better performance, and may use it to predict parameter values in the future.

As such, as a result of performing the remaining processes (S220-S250) of FIG. 2 only for some first combinations (first group) rather than all of them according to the amount of work at the time when the plurality of first combinations are extracted, the electronic device 100 There is an effect that the real-time workload of can be appropriately adjusted, and a target combination (ex. a target combination obtained from the first group) can be provided without waiting for a long time from the user's point of view. In particular, since a plurality of second combinations can be obtained and their performance compared for the remaining second groups, the effect of utilizing all of the plurality of first combinations extracted initially can be maintained.

Meanwhile, the electronic device 100 may determine whether to perform steps S220 to S250 with respect to the remaining groups (eg, the second group) according to the performance of the first target combination.

For example, the electronic device 100 may compare the performance of the first target combination with the average performance of all first combinations included in the first group.

If the performance of the first target combination is higher than the average performance by a threshold or more, the electronic device 100 may not perform processes S220 to S250 with respect to the second group (: do not obtain the second target combination). In this case, the first target combination is selected as the final target combination and can be used to predict parameter values in the future.

On the other hand, when the performance of the first combination is not higher than the average performance by a threshold value or more, the electronic device 100 may acquire the second target combination by performing processes S220 to S250 also for the second group.

In this case, the electronic device 100 compares the performance of the first target combination and the second target combination, selects one target combination having better performance, and can use it to predict parameter values later.

Here, when the performance of the first combination is superior to that of the second combination, the electronic device 100 may change and set the above-described threshold to be smaller. On the other hand, if the performance of the second combination is superior to the performance of the first combination, the electronic device 100 may maintain the above-described threshold value or change and set it to be greater.

Thereafter, in searching for a target combination by dividing a plurality of other parameters into a plurality of groups, the set threshold may be utilized as described above.

5 is a block diagram for explaining the configuration of an electronic device according to various embodiments of the present disclosure.

Referring to FIG. 5 , the electronic device 100 may include a communication unit 130, a user input unit 140, an output unit 150, and the like in addition to the memory 110 and the processor 120.

The communication unit 130 may include a circuit, module, chip, etc. for performing communication with at least one external device through various wired/wireless communication methods.

The communication unit 130 may be connected to external devices through various networks.

The network may be a Personal Area Network (PAN), a Local Area Network (LAN), a Wide Area Network (WAN), etc., depending on the area or size, and an intranet, It may be an extranet or the Internet.

The communication unit 130 performs long-term evolution (LTE), LTE Advance (LTE-A), 5th generation (5G) mobile communication, code division multiple access (CDMA), wideband CDMA (WCDMA), and universal mobile telecommunications system (UMTS). , WiBro (Wireless Broadband), GSM (Global System for Mobile Communications), DMA (Time Division Multiple Access), WiFi (Wi-Fi), WiFi Direct, Bluetooth, NFC (near field communication), Zigbee, etc. It can be connected with external devices through

In addition, the communication unit 120 may be connected to external devices through a wired communication method such as Ethernet, optical network, universal serial bus (USB), or thunderbolt.

The user input unit 140 is a component for receiving various commands or information from a user. The user input unit 140 may be implemented with at least one button, a touch pad, a touch screen, a microphone, a camera, a sensor, and the like. Also, the electronic device 100 may be connected to an input device for receiving user input, such as a mouse or keyboard.

For example, the electronic device 100 may perform an operation by loading an artificial intelligence algorithm for predicting a value of at least one parameter according to a user input received through the user input unit 140 .

The output unit 150 is a component for outputting various information and may include a display, a speaker, an earphone/headphone terminal, and the like. For example, the electronic device 100 may display the value of the predicted parameter on the display.

Meanwhile, the various embodiments described above may be implemented by combining two or more embodiments as long as they do not conflict or contradict each other.

Meanwhile, various embodiments described above may be implemented in a recording medium readable by a computer or a similar device using software, hardware, or a combination thereof.

According to the hardware implementation, the embodiments described in this disclosure are application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), and field programmable gate arrays (FPGAs). ), processors, controllers, micro-controllers, microprocessors, and electrical units for performing other functions.

In some cases, the embodiments described herein may be implemented by a processor itself. According to software implementation, embodiments such as procedures and functions described in this specification may be implemented as separate software modules. Each of the software modules described above may perform one or more functions and operations described herein.

On the other hand, computer instructions or computer programs for performing processing operations in the electronic device 100 according to various embodiments of the present disclosure described above are provided on a non-transitory computer-readable medium can be stored in Computer instructions or computer programs stored in such a non-transitory computer readable medium, when executed by a processor of a specific device, cause the above-described specific device to perform processing operations in the electronic device 100 according to various embodiments.

A non-transitory computer readable medium is a medium that stores data semi-permanently and is readable by a device, not a medium that stores data for a short moment, such as a register, cache, or memory. Specific examples of the non-transitory computer readable media may include CD, DVD, hard disk, Blu-ray disk, USB, memory card, ROM, and the like.

Although the preferred embodiments of the present disclosure have been shown and described above, the present disclosure is not limited to the specific embodiments described above, and is common in the technical field belonging to the present disclosure without departing from the gist of the present disclosure claimed in the claims. Of course, various modifications and implementations are possible by those with knowledge of, and these modifications should not be individually understood from the technical spirit or perspective of the present disclosure.

100: electronic device 110: memory
120: processor

Claims (7)

In the method for determining the direction prediction modeling method of a financial investment product by an electronic device,
Performing pre-processing on a plurality of parameters related to economic indicators;
extracting a plurality of first combinations including a first number of parameters randomly selected from among the plurality of parameters for which the preprocessing has been performed;
identifying performance of each of the plurality of first combinations by inputting values of parameters included in each of the plurality of first combinations to an artificial intelligence algorithm;
selecting at least one first combination having excellent performance from among the plurality of first combinations based on the identified performance;
extracting a plurality of second combinations obtained by dimensionally reducing the selected first combination; and
identifying performance of each of the plurality of second combinations by inputting values of parameters included in each of the plurality of second combinations into an artificial intelligence algorithm;
including,
Performing the preprocessing step,
Obtaining a ratio R _T of the economic indicator data of the current time to the economic indicator data of the previous time; and
The normalized S _T for the ratio R _T is expressed as

Steps to acquire according to
A modeling method determining method comprising a. According to claim 1,
The step of extracting a plurality of the first combinations,
Obtaining a predetermined number of first combinations including the first number of parameters randomly selected from among the plurality of parameters;
The preset number is,
The method of determining the modeling method, wherein the first number of parameters from the plurality of parameters is smaller than the number of all cases in which the parameters can be selected. According to claim 2,
The preset number is,
A method for determining a modeling method, which is set according to the number of the plurality of parameters. According to claim 1,
The step of extracting a plurality of second combinations,
Obtaining a plurality of second combinations that match the number of all cases in which a second number of parameters among the parameters included in the selected first combination can be selected,
The second number is less than the first number, modeling method determining method. According to claim 1,
Identifying the performance of each of the plurality of first combinations,
Identifying performance of each of the plurality of first combinations using an artificial intelligence algorithm consisting of at least one of random forest, extreme gradient boosting (XGBoost), K-nearest neighbor (KNN), and support vector machine (SVM),
Identifying the performance of each of the plurality of second combinations,
A modeling method of identifying the performance of each of the plurality of second combinations by using an artificial intelligence algorithm composed of at least one of a random forest, XGBoost (Extreme Gradient Boosting), KNN (K-Nearest Neighbor), and SVM (Support Vector Machine) How to decide. According to claim 1,
The method for determining the modeling method,
selecting at least one second combination having excellent performance among the plurality of second combinations as a target combination; and
Including, the step of predicting the value of at least one parameter by inputting the values of the parameters constituting the target combination to the artificial intelligence algorithm. According to claim 6,
The method for determining the modeling method,
extracting a plurality of first combinations including a first number of parameters randomly selected from among the plurality of parameters when the accuracy of the predicted parameter value is less than the threshold value;
further selecting at least one target combination based on the re-extracted plurality of first combinations; and
Comparing the performance of the selected target combination and the additionally selected target combination; including, modeling method determining method.

Images