JP2011227794A - Site analysis system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a site analysis system which is capable of generating a portfolio about a plurality of sites to generate reports allowing states of the sites to be recognized per event of the portfolio.SOLUTION: In the site analysis system, model generation means 11 takes a plurality of sites as analysis target sites based on map information, statistical information, and site attribute information and performs multiple regression analysis processing with respect to the analysis target sites to generate a calculation formula of multiple repression analysis as an analytic model, and existing-site analysis means 12 generates a portfolio based on the analytic model and ranks the sites by events of the portfolio to generate a report including an action plan, and new-site analysis means 13 uses the analytic model to generate a report about a new site.

Description

  The present invention relates to a system for analyzing bases such as stores, hospitals, factories, and the like, and more particularly to a base analysis system that generates a portfolio by analyzing each of a plurality of bases and creates a report for each event of the portfolio.

[Conventional technology]
Some conventional trade area analysis systems analyze sales and profits at bases such as stores.

[Related technologies]
As related prior art documents, Japanese Patent Application Laid-Open No. 2003-167880 “Statistical Information Aggregation Device” (Kao Corporation) [Patent Document 1], Japanese Patent Application Laid-Open No. 2003-345996 “Store Information Collection and Supply Device, Store Information” Collection and providing method and program describing it "(NC System Technology Co., Ltd.) [Patent Document 2].

  Patent Document 1 is a device that designates a store on a map and displays data such as the sales amount of the store together with map information, acquires data held by the POS system, creates a display file, It is described that when data to be displayed in a file is designated, it is displayed on a map to support analysis work.

  Patent Document 2 is a device that evaluates the location of a store and provides an analysis result, receives store location information from an information provider, transmits store information to the user, and receives response information from the user. It is described that an evaluation of a user corresponding to a store position is transmitted to a person who wants to obtain information, and the user's reaction behavior is digitized and aggregated based on response information from the user, and the evaluation of the user with respect to the store position is calculated. .

JP 2003-167880 A JP 2003-345932 A

  However, although the above conventional analysis system analyzes individual stores, etc., it has a function of generating a report that groups characteristics of a plurality of stores and makes it easy to grasp the situation of the base. There was no problem.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a site analysis system capable of generating a portfolio for a plurality of sites and creating a report that enables the status of the sites to be grasped for each event of the portfolio. And

  The present invention for solving the problems of the conventional example described above is a site analysis system in which a control unit that reads a processing program and executes processing, a map information storage unit that stores map information, and a statistics that stores statistical information An information storage unit and a site attribute information storage unit for storing site attribute information are provided. As processing means realized by the control unit, map information is read from the map information storage unit, and statistical information corresponding to the map information is statistically calculated. Read from the information storage unit, read the base corresponding to the map information from the base attribute information storage unit, narrow down multiple bases according to conditions for the base attribute information, extract the bases to be analyzed, and perform multiple regression analysis processing on the bases to be analyzed To create a calculation formula for multiple regression analysis as an analysis model and calculate the predicted value of each site using the analysis model; Using the actual value and the predicted value included in the base attribute information of the information storage unit, and an existing base analyzing means for generating a portfolio having four events divided by vertical and horizontal lines for the analysis model And

  The present invention includes an action plan storage unit that stores action plan information for each event of a portfolio for each analysis model in the site analysis system, and the existing site analysis unit uses an event for an analysis model created by the model creation unit. It is characterized by reading an action plan every time and creating a report.

  According to the present invention, in the above-described site analysis system, when the existing site analysis means does not store an action plan in the action plan storage unit for a new model, an approximate analysis model is searched, and an action of the searched analysis model It is characterized by reading a plan.

  The present invention is characterized in that, in the site analysis system, the existing site analysis means ranks sites for each event.

  The present invention is characterized in that, in the base analysis system, the existing base analysis means includes portfolio adjustment means for adjusting the number of bases for each event by shifting a vertical line and a horizontal line for dividing four events.

  According to the present invention, in the above-described site analysis system, the model creation means selects a variable having a high multicollinearity, a normalization process that specifies a unit of the input variable and a normalization type before the multiple regression analysis process. It is characterized by performing a multicollinearity check process to be excluded.

  In the site analysis system, the present invention provides a principal component analysis process in which the model creating means aggregates variables between the multicollinearity check process and the multiple regression analysis process or / and a cluster analysis process in which data grouping is performed. It is characterized by performing.

  In the multi-collinearity check process according to the present invention, the multi-collinearity check process performs a correlation calculation of variables between the bases, and performs a process of excluding those having a high correlation coefficient or the absolute value of the correlation coefficient. Features.

  The present invention uses the analysis model created by the model creation means in the site analysis system, artificially sets site attribute information related to the site from real estate information, and further uses statistical information and / or other information used in the analysis model By adding the base attribute information of the bases, a predicted value when a base is newly established is calculated, and a new base analysis means for generating a report regarding the base is provided.

  According to the present invention, in the site analysis system, the site attribute information includes address information of a plurality of customers in addition to the location information of the sites, and the model creating means includes the location information of the sites and the plurality of customers. Find the distance between the base and multiple customers from the address information or the required time of the distance, find the actual trade area based on the average distance averaged distance or average time required, and map the actual trade area It displays on the map information read from the information storage part.

  According to the present invention, in the site analysis system, the site attribute information includes address information of a plurality of customers in addition to the location information of the sites, and the model creating means includes the location information of the sites and the plurality of customers. Find the distance between the base and multiple customers from the address information or the required time of the distance, and the mode, maximum value, minimum value, variance, quartile value, or outlier of the distance or required time A real trade area is obtained based on the index, and the real trade area is displayed on the map information read from the map information storage unit.

  According to the present invention, in the site analysis system, the model creation means can display a histogram or a box-and-whisker plot for the distance between the site and a plurality of customers or the time required for the distance, separately from the display of the map information. It is characterized by doing.

  According to the present invention, the model creation means reads the map information from the map information storage unit, reads the statistical information corresponding to the map information from the statistical information storage unit, and determines the base corresponding to the map information from the base attribute information storage unit. Read and base attribute information to narrow down multiple bases according to conditions and make them analysis target bases, perform multiple regression analysis processing on the target bases to generate multiple regression analysis formulas as analysis models, The existing base analysis means uses the actual value and the predicted value included in the base attribute information of the base attribute information storage unit, and uses the actual value and the predicted value to analyze the four events divided into vertical and horizontal lines. Because it is a site analysis system that generates a portfolio that has There is an effect that can create a report that can grasp the situation of the point.

It is a lineblock diagram of a base analysis system concerning an embodiment of the invention. It is a block diagram of a process means and a memory | storage part. It is the schematic for producing | generating an analysis model. It is a flowchart of a model creation process. It is a flowchart of the existing base analysis process. It is an example of the display screen which shows the outline of a portfolio. It is a figure explaining real trade area logic. It is a figure which shows the example of a map display of a base.

Embodiments of the present invention will be described with reference to the drawings.
[Outline of the embodiment]
In the site analysis system according to the embodiment of the present invention, the model creation means extracts a plurality of sites as analysis target sites based on map information, statistical information, and site attribute information, and each of the extracted analysis target sites Perform multiple regression analysis processing to create multiple regression analysis formulas as analysis models, and the existing site analysis means create a portfolio for the analysis model, and create an action plan by ranking sites by portfolio event Therefore, it is possible to grasp the status of existing bases for each portfolio event and report an appropriate action plan.

[Overall configuration of this system: Fig. 1]
A site analysis system according to an embodiment of the present invention will be described with reference to FIG. FIG. 1 is a configuration diagram of a site analysis system according to an embodiment of the present invention.
As shown in FIG. 1, the site analysis system (this system) according to the embodiment of the present invention is configured by a computer, and includes a control unit 1, a storage unit 2, and an interface unit 3, and includes an interface unit. 3, a display device 4, a printing device 5, and an input device 6 that are peripheral devices are connected to the network.

The configuration of FIG. 1 assumes a single computer, but includes a server that is a computer that executes processing, and a client (terminal) that is a computer connected to the server via a network, and an input unit of the client The server may be accessed by an instruction from, and the processing result at the server may be displayed on the display unit of the client.
Alternatively, the Internet may be used as a network, and the server may be a Web server.

[Parts of this system: Fig. 2]
Each part of the system will be described with reference to FIGS. FIG. 2 is a configuration diagram of the processing means and the storage unit.
[Storage unit 2]
The storage unit 2 includes a map information storage unit 2a, a statistical information storage unit 2b, a base attribute information storage unit 2c, and an action information storage unit 2d. The storage unit 2 also stores a processing program.

The map information storage unit 2a stores map information (map data).
The statistical information storage unit 2b stores various statistical information (statistical data).
The base attribute information storage unit 2c stores information related to base attributes (regional information such as location information such as an address, actual data related to the base). POS (Point Of Sale) information (including actual values such as sales and profits), medical equipment and patient information for a hospital, factory equipment information for a factory, and the like are stored.

The action information storage unit 2d stores action information (correspondence information) for each event of the portfolio for each analysis model, and is used for report creation.
The action information storage unit 2d is a learning type knowledge DB. The learning knowledge DB stores reports created by users for each industry, and when the user creates a report for a new analysis model, the corresponding action plan is not stored for each industry. An analysis model to be approximated is searched, and a report of the analysis model can be referred to as a sample.

[Display device 4, printing device 5, input device 6]
The display device 4 displays a screen for performing analysis model creation, site analysis, and the like, which will be described later, and further displays an analysis model, analysis results, and a report, and can be printed by the printing device 5.
Further, the input device 6 inputs instructions for creating an analysis model, analyzing a site, and the like.

[Control unit 1]
The control unit 1 reads a processing program stored in the storage unit 2 and realizes an analysis model creation unit 11, an existing site analysis unit 12, and a new site analysis unit 13.
Hereinafter, each processing means will be described.

[Model creation means 11]
The model creation means 11 includes map information stored in the map information storage unit 2a, statistical information stored in the statistical information storage unit 2b, and base attribute information stored in the base attribute information storage unit 2c (if it is a store, information on regional characteristics of the store, sales / profit An analysis model is created using information (actual values).

  Specifically, the model creating means 11 specifies a plurality of bases in the specific area using the map information of the area specified by the user, the statistical information specified by the user, and the position information in the area information of the base attribute information, Narrow down the bases by conditions for the base attribute information, extract the bases to be analyzed, perform multiple regression analysis processing on each extracted base, create a formula for multiple regression analysis as an analysis model, and use the analysis model to determine the potential of each base (A predicted sales value if the site is a store) is calculated, and the brute force correlation of each site is taken with respect to the site attribute information, and those that are approximated are excluded and set as the final analysis target site.

  That is, the model creation means 11 extracts a plurality of elements from the site attribute information and performs narrowing down and analysis. Specifically, normalization processing, multicollinearity check processing, and multiple regression analysis processing are performed. Note that the principal component analysis process and the cluster analysis process may be performed after the multicollinearity check process.

[Normalization processing]
In the normalization process, as a preparation for the input variable, a normalization method is selected by automatic discrimination or visual confirmation. Specifically, the unit of the input variable and the type of normalization are specified.
The unit of the input variable is money amount, number of people, area, volume, efficiency, etc., and the type of normalization is selected according to the unit and the number of digits. The type of normalization selected is real number, logarithm, reciprocal number, addition + log, ratio, and the like.

In the normalization process, the missing value processing method is selected by automatic discrimination or by providing candidates. Specifically, the missing values are discarded, set to 0, averaged, or 10 or 100 are added to the missing values for log.
In normalization processing, indices such as mean, standard deviation, variance, distortion, sharpness, variation coefficient, and graphs such as histograms and box-and-whisker charts can be used.

[Multi-collinearity check processing]
The multi-collinearity check process is for selecting and excluding variables with high multi-collinearity. Variables are automatically selected, for example, by cutting off the correlation coefficient or the absolute value of the correlation coefficient from the highest. You can choose. Here, the correlation coefficient is a correlation value obtained by performing a brute force correlation calculation using numerical data as variables for the site attribute information for each site.
The multicollinearity check process can also use a VIF (Variance Inflation Factor) index.

[Principal component analysis]
The principal component analysis process can be expected to collect variables and suppress multicollinearity. For example, 50 variables are aggregated into 15.
Specifically, aggregation can be achieved by automatically extracting factors having an eigenvalue of 1 or more and not using variables less than 1.

  Note that factor analysis may be performed instead of principal component analysis. Principal component analysis is intended to construct a composite score from observed data, whereas factor analysis is intended to obtain individual components assuming that the observed data is a composite quantity. .

[Multiple regression analysis processing]
The multiple regression analysis process is for creating a prediction model, is a multivariate analysis, and has multiple independent variables for regression analysis. By selecting multiple appropriate variables, it is possible to create a prediction formula (calculation formula for multiple regression analysis) that is easy to calculate and has few errors. Commonly used are least squares and multiple regression of linear models.

In addition, multiple regression analysis can suppress multicollinearity by a stepwise method.
Specifically, the model creation unit 11 calculates a predicted value (a sales predicted value in the case of a store) by multiple regression analysis processing.

[Cluster analysis processing]
The model creation means 11 may perform grouping by cluster analysis before creating a portfolio.
The cluster analysis process is a method of automatically classifying given data without external criteria, and here, a hierarchical method in which data is classified hierarchically is used.

  Moreover, you may utilize cross segment cluster analysis as cluster analysis. Cross-segment cluster analysis is a cross-tabulation that uses segments newly identified in cluster analysis as a tabulation axis, and this makes it possible to understand clusters more widely and deeply.

[Analysis model: Fig. 3]
A method of obtaining an analysis model, that is, a calculation formula for multiple regression analysis will be described with reference to FIG. FIG. 3 is a schematic diagram for generating an analysis model.
The calculation formula of the multiple regression analysis generated by the multiple regression analysis process is simply expressed by the following formula, for example.
y = b0 + b1 .x1 + b2 .x2 + b3 .x3
Here, y indicates a potential such as sales, b0 to b3 are coefficients, and x1 to x3 are elements in the base attribute information used in the calculation formula of the multiple regression analysis.
Therefore, the calculation formula (analysis model) of the multiple regression analysis differs depending on how the elements in the site attribute information are selected.

In FIG. 3, for example, when the base is a store, for stores A to Z, element x1 is the number of employees, element x2 is the number of parking lots, and element x3 is the number of competing stores. .
Here, the elements on the horizontal axis can be aggregated or the elements on the vertical axis can be grouped. The elements on the horizontal axis can be aggregated by principal component analysis, and the stores on the vertical axis can be grouped. Is the cluster analysis.

[Existing site analysis means 12]
The existing site analysis unit 12 generates a portfolio of the actual value in the site attribute information and the predicted value obtained by the multiple regression analysis process of the model creation unit 11. In the display screen, the vertical axis is the actual value, the horizontal axis is the predicted value (potential), and a plurality of bases are displayed as dots on the screen. Furthermore, a plurality of locations on the screen are classified into four events by horizontal lines and vertical lines.

Furthermore, the existing site analysis means 12 ranks sites according to portfolio events, reads action information for each event from the action storage unit 2d, and creates an action plan as a report. Here, the ranking is a higher rank (a thing that must be dealt with urgently) as it is farther from the intersection of the vertical and horizontal lines.
When the action information of the new model is not stored in the action storage unit 2d for the new model, the existing site analysis unit 12 searches for an approximate (similar) model, and the action information of the searched model is obtained. It can be referred to.

Even if the portfolio is not created, the overall ranking may be performed by making a Z score.
Further, the existing site analysis means 12 may generate a portfolio in a group grouped by cluster analysis, and perform ranking for each event and creation of an action plan.

The existing base analysis means includes portfolio adjustment means.
[Portfolio adjustment means]
The portfolio adjustment means adjusts the number of bases for each event by scrolling and moving (shifting) the vertical and horizontal positions of the created portfolio.
The portfolio adjustment unit may automatically adjust the vertical line and the horizontal line according to the condition setting.

[New site analysis means 13]
The new site analysis unit 13 uses the analysis model (calculation formula of multiple regression analysis) created by the model creation unit 11 to set information on the site in a pseudo manner from the real estate information, and also simulates the value of the site attribute information. By setting it manually, an evaluation is performed when a new base is established. Here, the evaluation is to create a report by performing a calculation using a calculation formula (analysis model) of multiple regression analysis for a newly planned base.

  That is, in the example of FIG. 3, the new site analysis means 13 sets element x1 (number of employees) as an assumed value in the analysis model (calculation formula of multiple regression analysis) created by the model creation means 11, and element x2 (number of employees) The predicted number of sales (sales potential) can be obtained by setting the number of parked vehicles from real estate information and setting element x3 (number of competing stores) from the site attribute information.

Next, the processing flow of this system will be described.
[Model creation process: Fig. 4]
The model creation process in the model creation unit 11 will be described with reference to FIG. FIG. 4 is a flowchart of the model creation process.
First, as shown in FIG. 4, the model creation means 11 reads map information of a region designated by the user from the map information storage unit 2a (S1), and further, statistical information specified by the user for statistical information related to the region. Is read from the statistical information storage unit 2b (S2).

  Next, the model creation means 11 narrows down the bases based on the conditions using the base attribute information in the base attribute information storage unit 2c (S3). Specifically, using the regional information (location data such as address) of the base attribute information, the base corresponding to the region of the read map information is specified, and further, the base attribute information is selected according to the condition, and the base is narrowed down. Do.

  Then, a plurality of narrowed bases are set as analysis target bases (S4), the above-described normalization process is performed (S5), and then a multicollinearity check process is performed (S6). Further, it is determined whether or not execution of principal component analysis processing is set (S7). If it is set (in the case of Yes), principal component analysis processing is performed (S8). If principal component analysis execution is not set (in the case of No), the process proceeds to step S9.

  Next, it is determined whether or not to execute cluster analysis processing (S9). If it is set (in the case of Yes), cluster analysis processing is performed (S10). If cluster analysis execution is not set (in the case of No), the process proceeds to step S11.

Then, a multiple regression analysis process (S11) is performed, and a calculation formula for multiple regression analysis for the analysis target site is created as an analysis model.
In addition, the predicted value of the base obtained by the multiple regression analysis process is output to the existing base analysis means 12 and used for the existing base analysis process.

[Existing site analysis processing: Fig. 5]
The existing site analysis process in the existing site analysis means 12 will be described with reference to FIG. FIG. 4 is a flowchart of the existing site analysis process.
As shown in FIG. 5, the existing site analysis unit 12 generates a portfolio based on the predicted value obtained in the multiple regression analysis process of the model creation process and the actual value stored in the site attribute information storage unit 2 c ( S21).

Next, the existing site analysis means 12 ranks sites according to portfolio events (S22), reads action information corresponding to or approximates each event from the action storage unit 2d (S23), and creates an action plan report. (S24). For the report, the action information in the action storage unit 2d may be used as it is, or the user may add and modify the action information to complete the report.
The created report is stored in the action storage unit 2d.

Next, the display screen will be described.
[Portfolio display screen: Fig. 6]
The outline of the portfolio will be described with reference to FIG. FIG. 6 is an example of a display screen showing an outline of the portfolio.
As shown in FIG. 6, the outline of the portfolio is the actual value on the vertical axis and the potential value (predicted value) on the horizontal axis. In FIG. 6, the actual value and the potential value are used on the vertical and horizontal axes. However, the actual values and the potential values may be used on the vertical and horizontal axes.
The display screen is classified into vertical lines and horizontal lines. Event A is in the upper right, Event B is in the upper left, Event C is in the lower right, and Event D is in the lower left.

Hereinafter, for each event, the characteristics when the base is a store will be described.
[Event A]
The store of event A is a flower-shaped store with a high actual value (profit) and a high potential value.
The point in the report is, for example, “At a store with high marketability and stable profits, aiming to maintain the current situation without incurring costs”.

  Also, for example, the action plan commented, “Hanagata store with high marketability and high profits. It is a royal store where customers can gather without investing, and profits are invested in other stores. “We don't invest aggressively because we are making profits that match our potential, so we focus more on maintaining the status quo rather than aiming for further growth.”

[Event B]
The store of event B is a store of trees made of gold having a high actual value (profit) and a low potential value.
The point in the report is, for example, “Aiming to maintain the current situation while making a minimum investment in an efficient and efficient store with low marketability but high profits”.

  In addition, for example, the action plan commented, “Excellent stores that increase profits relative to the market size. Many stores have good store operations and maximize the potential of the stores. “Because of the low nature of the store, the store lifecycle tends to decline, so a minimum investment is needed to stop it, and a branding strategy is also effective to add high value.”

[Event C]
The store of event B is a problem store with a low actual value (profit) and a high potential value.
The point in the report is, for example, “A store with a problem that has high marketability but does not increase profits, and has high potential, so we will aggressively invest and restructure to achieve high growth”.

  The action plan commentary is, for example, “Stores with high potential but not profitable. Many stores have problems with the internal factors of the store. Because it is expensive and can expect significant growth, it is necessary to take drastic levers such as promotion, store renovation, review of store operations, and personnel reforms. Sometimes. ”

[Event D]
The store of event B is a withdrawal considering store with a low actual value (profit) and a low potential value.
The point in the report is, for example, “Stores with low marketability and profits, which are not expected to have an effective cost-effectiveness for investment, and are considering withdrawal”.
An explanation of the action plan is, for example, “Stores that are low in marketability and profits and cannot be expected to have a large effect even if they are invested.

[Option (applied technology)]
Next, the contents of actual commercial area logic and map display will be described as applied technologies.
[Real business area logic: Fig. 7]
When the variable is created by the model creating means 11, it is a logic for creating a variable that serves as an index for the concept of the actual commercial area of the facility.
If the base is a store, the distance between the store and the customer who visits the store (the distance connecting the store's address point and the customer's address point) or the time required for each customer to visit the store is calculated. The average distance or average time is calculated from the time.
The value is used as an index for the actual trade area and displayed on the map to grasp the actual trade area.

In this case, instead of simply calculating the average, the index, such as average, standard deviation, variance, distortion, sharpness, coefficient of variation, etc., can be expressed numerically, or by graphs such as histograms and box plots. You may make it grasp.
Specifically, for example, as shown in FIG. 7, the distance between the customers A to E and the store and the required time are obtained, and the distance or time is totaled to obtain the average. For example, the circle of trade area drawn around the store is drawn with the average distance as the radius. In addition, a walking area, a car or a train trade area is drawn from the average time, and is a diagram of an indefinite area (area surrounded by a broken line). In other words, in the case of time, a distorted area is formed depending on the landform and road conditions.
FIG. 7 is a diagram for explaining the actual commercial area logic.

[Map display of base: Fig. 8]
The analyzed event is displayed as an icon by the portfolio based on multiple regression analysis, and the circle or circle based on the average distance obtained by the actual market area logic, the walk by the average time, or the car or train trade area is superimposed on the store on the map. You may make it display.
FIG. 8 is a diagram showing a map display example of the base. FIG. 8 shows an example in which the type of event is entered at the center of the base, and the circle trade area based on the average distance is drawn.
In addition to drawing the average distance and average time, drawing indices such as the mode, maximum value, minimum value, variance, quartile, and outlier on the map based on the distance or time (distance In the case of (3), a circular area may be drawn, and in the case of time, an irregular area may be drawn), or a histogram or box-and-whisker chart may be displayed in a separate frame from the map for reference.

[Effect of the embodiment]
According to this system, the model creating means 11 performs multiple regression analysis processing on each of a plurality of bases based on map information, statistical information, and base attribute information to create an analysis model (multiple regression analysis calculation formula). The existing base analysis means 12 creates a portfolio for the analysis model, and ranks the bases according to the portfolio events so as to create an action plan. It has the effect of reporting an appropriate action plan.

  Further, according to the present system, the new site analysis means 13 can perform an evaluation when a new site is established based on the created analysis model, and it is possible to easily determine whether or not a new site can be established.

  INDUSTRIAL APPLICABILITY The present invention is suitable for a site analysis system that can generate a portfolio for a plurality of sites and create a report that can grasp the status of the site for each event of the portfolio.

  DESCRIPTION OF SYMBOLS 1 ... Control part, 2 ... Storage part, 2a ... Map information storage part, 2b ... Statistical information storage part, 2c ... Base attribute information storage part, 2d ... Action information storage part, 3 ... Interface part, 4 ... Display apparatus, 5 ... Printing device, 6 ... Input device, 11 ... Analysis model creation means, 12 ... Existing site analysis means, 13 ... New site analysis means

Claims (12)

A control unit that reads a processing program and executes processing, a map information storage unit that stores map information, a statistical information storage unit that stores statistical information, and a base attribute information storage unit that stores base attribute information,
As processing means realized by the control unit,
Read the map information from the map information storage unit, read the statistical information corresponding to the map information from the statistical information storage unit, read the base corresponding to the map information from the base attribute information storage unit, about the base attribute information Narrow down multiple bases according to conditions, extract analysis target bases, perform multiple regression analysis processing on the analysis target bases, generate multiple regression analysis formulas as analysis models, and use the analysis models to calculate predicted values for each base A model creation means for calculating;
Using the actual value and the predicted value included in the site attribute information of the site attribute information storage unit, the existing site analysis means for generating a portfolio having four events divided by vertical and horizontal lines for the analysis model; A site analysis system characterized by having. With an action plan storage unit that stores action plan information for each portfolio event for each analysis model,
2. The site analysis system according to claim 1, wherein the existing site analysis means reads an action plan for each event and creates a report for the analysis model created by the model creation means.   The existing base analysis means, when no action plan is stored in the action plan storage unit for a new model, searches for an approximate analysis model and reads an action plan of the searched analysis model. 2. Base analysis system according to 2.   4. The site analysis system according to claim 1, wherein the existing site analysis means ranks sites for each event.   The base analysis according to any one of claims 1 to 4, wherein the existing base analysis means comprises portfolio adjustment means for adjusting the number of bases for each event by shifting the vertical and horizontal lines that divide the four events. system.   Before the multiple regression analysis process, the model creation means includes a normalization process that identifies the unit of the input variable and the type of normalization, and a multi-commonness check process that selects and excludes variables having high multicollinearity. 6. The site analysis system according to claim 1, wherein the site analysis system is performed.   The model creating means performs a principal component analysis process for aggregating variables and / or a cluster analysis process for grouping data between the multiple co-glandularity check process and the multiple regression analysis process. 6. Site analysis system according to 6.   8. The multiple co-glandularity check process performs a correlation operation of variables between bases, and performs a process of excluding a correlation coefficient or a coefficient having a high absolute value of a correlation coefficient. Site analysis system.   Using the analysis model created by the model creation means, the base attribute information related to the base is set in a pseudo manner from the real estate information, and the statistical information used in the analysis model and / or the base attribute information of other bases are aggregated. 9. The site analysis system according to claim 1, further comprising a new site analysis unit that calculates a predicted value when a site is newly established and generates a report related to the site. In addition to the location information of the location, the location attribute information includes address information of multiple customers.
The model creating means obtains the distance between the base and the plurality of customers or the time required for the distance from the location information of the base and the address information of the plurality of customers, and averages the distance or the average distance or The base analysis system according to any one of claims 6 to 9, wherein an actual commercial area is obtained based on an average time obtained by averaging the required times, and the actual commercial area is displayed on the map information read from the map information storage unit. . In addition to the location information of the location, the location attribute information includes address information of multiple customers.
The model creation means obtains the distance between the base and the plurality of customers or the required time of the distance from the location information of the base and the address information of the plurality of customers, and the mode value for the distance or the required time A real trade area is obtained based on an index of maximum value, minimum value, variance, quartile value, or outlier, and the real trade area is displayed on the map information read from the map information storage unit. Item 10. The site analysis system according to any one of Items 6 to 9.   The model creating means can display a histogram or box-and-whisker plot for the distance between the base and the plurality of customers or the time required for the distance, separately from the display of the map information. 11. The site analysis system according to 11.

Images