JP2013131052A - Information processing device, control method therefor, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the accuracy of predicting prediction data on an inheritance destination item.SOLUTION: In accordance with a combination of result data of an inheritance source item and result data of an inheritance destination item which are stored, prediction data on the inheritance destination and inheritance source items is calculated. In accordance with the result data of the inheritance source item which is stored, prediction data on the inheritance source item is calculated. By subtracting the prediction data on the inheritance source item from the prediction data on the inheritance destination and inheritance source items, prediction data on the inheritance destination item is calculated and output.

Description

  The present invention relates to an information processing apparatus, a control method thereof, and a program, and more particularly to a technique for predicting prediction data of an inheritance item.

  Conventionally, when forecasting with a time series forecasting model, etc., based on the specified number of past time series data reference periods, time series data for the reference period is obtained and forecasted based on the actual data for the reference period The amount is calculated (Non-Patent Document 1).

  In addition, it is a common story that products are renewed for reasons such as changing packages or increasing the amount by 10%. At that time, although the product before renewal and the product after renewal are the same for consumers, the product codes are different for production or sales, so they are managed as different products in the management of performance data. The In this case, the product after renewal is predicted as a new product.

  As a method for predicting the sales degree of such a new product, for example, Patent Document 1 discloses that when predicting the sales degree of a new product, sales data of a newly released product and sales data of other existing products It is proposed that an existing product having the highest correlation is selected and the sales degree of the newly released product is predicted from the sales data of the selected existing product thereafter.

“Nothing to predict” by Taira Omura, Nikka Giren JP 2002-169946 A

  However, even if the conventional technology is used, in the case of a new product with little information (actual value) used for prediction processing such as sales data of existing products different from the new product, it is possible to obtain the predicted value of the new product with high accuracy. difficult.

  Therefore, for example, when predicting a product after a renewal as a new product (inherited item), the product before the renewal (inheriting source item) is regarded as the same product as the product after the renewal, and before the renewal. By using the actual value of the product, it is possible to compensate for the insufficient actual value in the prediction process of the product after the renewal.

  Therefore, the object of the present invention is to predict the inheritance destination item by subtracting the prediction data of the inheritance source item obtained according to the performance data of the inheritance source item from the prediction data obtained according to the performance data of the inheritance source item and the succession destination item. By calculating the data, the prediction accuracy of the prediction data of the inheritance item is improved.

  The present invention relates to storage means for storing the result data of the inheritance source item and the result data of the inheritance destination item, the achievement data of the inheritance source item stored in the storage means, and the inheritance destination item. First inheritance means for calculating predicted data for the inheritance destination and the inheritance source item according to a combination with the actual performance data, and the inheritance according to the actual data of the inheritance source item stored in the storage means By subtracting the prediction data calculated by the second calculation means from the prediction data calculated by the second calculation means for calculating the prediction data for the original item and the first calculation means, the inheritance Third calculation means for calculating prediction data for the previous item; output means for outputting prediction data calculated by the third calculation means; Characterized in that it comprises.

  The present invention is also a method for controlling an information processing apparatus comprising storage means for storing the result data of an inheritance source item and the result data of an inheritance destination item, the first calculation of the information processing apparatus Means for calculating prediction data for the inheritance destination and the inheritance source item according to a combination of the result data of the inheritance source item stored in the storage means and the performance data of the inheritance destination item; And a second calculation unit of the information processing apparatus calculates second prediction data for the inheritance source item according to the result data of the inheritance source item stored in the storage unit. The calculation step and the third calculation means of the information processing apparatus subtract the prediction data calculated in the second calculation step from the prediction data calculated in the first calculation step. A third calculation step of calculating prediction data for the inherited item, and an output step of outputting the prediction data calculated by the third calculation step by the output means of the information processing apparatus; It is characterized by providing.

  In addition, the present invention is a program that can be read and executed by an information processing apparatus including a storage unit that stores the result data of an inheritance source item and the result data of an inheritance destination item, and the information processing apparatus includes: A first calculation for calculating prediction data for the inheritance destination and the inheritance source item according to a combination of the performance data of the inheritance source item and the performance data of the inheritance destination item stored in the storage unit Means, second calculation means for calculating prediction data for the inheritance source item according to the result data of the inheritance source item stored in the storage means, and the prediction calculated by the first calculation means Subtracting the prediction data calculated by the second calculation means from the data to calculate prediction data for the inherited item A calculation means, characterized in that to function as output means for outputting the third predicted data calculated by the calculation means.

  According to the present invention, the prediction data of the inheritance destination item is subtracted from the prediction data obtained according to the performance data of the inheritance source item and the succession destination item, by subtracting the prediction data of the inheritance source item obtained according to the performance data of the inheritance source item. By calculating, the prediction accuracy of the prediction data of the inheritance item can be improved.

It is a lineblock diagram of information processor 100 concerning an embodiment of the present invention. It is a block diagram which shows the hardware constitutions of the information processing apparatus applicable to the information processing apparatus 100 shown in FIG. FIG. 3 is a diagram illustrating an example of a flowchart for performing control processing in the information processing apparatus 100. It is a flowchart which shows the detailed process of the prediction process of step S303 of FIG. It is a figure which shows the inheritance relationship between each item. It is an example of an inheritance setting list screen. It is an example of a parameter setting screen. It is an example of monthly sales amount data (time series data obtained by totaling actual values in time series) of the inheritance source item A002 and the succession destination item A201 for 36 months from January 2008 to December 2010. 9 is a graph of the time series data of FIG. It is an example of the graph which represented the correction | amendment performance value of item A201 of an inheritance destination, and the prediction result using the correction | amendment performance value in time series. It is a graph of the prediction result of inheritance origin item A002. It is an example of the graph which represented the past performance (result value of sales) of A201 which is an inheritance destination item, and the prediction result in time series. It is an example of the flowchart of the detailed process of step S407 of FIG. It is an example of the graph which represented the correction factor of each time from the correction start time (November 2010) to the correction end time (May 2012) obtained from each prediction correction model (calculation formula) in time series. A graph showing in time series the actual value of the inheritance source item A002, the predicted value calculated from the actual value, and the predicted value (corrected predicted value) obtained by correcting the predicted value by the prediction correction model It is an example. It is an example of the graph which represented the predicted value of inheritance destination item A201 in time series. It is an example of the flowchart of the detailed process of step S407 of FIG. Inheritance source item A002, actual results (actual values) in December 2010, forecasts after January 2011 (predicted values), inventory at the end of December 2010 (inventory number), and data after January 2011 It is an example of the data table which represented the warehousing schedule (scheduled warehousing value) in time series. The actual value of December 2010 of the inheritance item A002, the predicted value after January 2011, the stock quantity (actual value) at the end of December 2010, the planned warehousing value after January 2011, 2011 It is an example of the data table which represented the prediction upper limit after January, 2011, and the correction | amendment prediction value after 2011 yen January in time series. The corrected predicted value is represented by a graph. It is the graph which plotted the predicted value of inheritance destination item A201 obtained as a result of prediction of inheritance destination item A201 in time series.

  Hereinafter, with reference to the drawings, an information processing apparatus according to an embodiment of the present invention and details of its operation processing will be described.

  FIG. 1 is a configuration diagram of an information processing apparatus 100 according to an embodiment of the present invention.

  As illustrated in FIG. 1, the information processing apparatus 100 according to the present embodiment stores an inheritance information DB 104, a prediction model DB 105, a performance DB 106, and a prediction result DB 107 in an external memory 211. A determination unit 108, a prediction unit 102, and a display unit 103 are provided. The external memory 211 is an application example of the storage means of the present invention.

  The inheritance information DB 104 stores information on contents set in FIG. 6 to be described later. The prediction model DB 105 stores information on contents set in FIG.

  The result DB 106 stores time series data of FIG. 8 to be described later. That is, the result DB 106 stores the result data of the inheritance source item and the result data of the inheritance destination item.

  The prediction result DB 107 stores a prediction value obtained as a result of prediction by the prediction unit 102.

  The setting unit 101 is a setting unit that sets various types of information input through user operations through FIGS. 6 and 7 described later, and stores them in the inheritance information DB 104 and the prediction model DB 105.

The order determination means 108 determines the order in which the prediction process predicted by the prediction means 102 is performed for each item according to the inheritance information stored in the inheritance information DB 104.

  Further, the prediction unit 102 uses various past prediction values (for example, actual value of sales data) and various predictable prediction models (for example, moving average, variable index) stored in the prediction model DB 105. This is a prediction unit that calculates a future predicted value (for example, a predicted value of future sales data) using a smoothing formula, a calculation formula such as winters without fluctuation, or the like.

  The display unit 103 is a display unit that displays a result (predicted value) predicted by the prediction unit 102.

  Hereinafter, the hardware configuration of the information processing apparatus applicable to the information processing apparatus 100 illustrated in FIG. 1 will be described with reference to FIG.

  FIG. 2 is a block diagram illustrating a hardware configuration of an information processing apparatus applicable to the information processing apparatus 100 illustrated in FIG.

  In FIG. 2, reference numeral 201 denotes a CPU that comprehensively controls each device and controller connected to the system bus 204. Further, the ROM 202 or the external memory 211 is necessary to realize a BIOS (Basic Input / Output System) or an operating system program (hereinafter referred to as an OS), which is a control program of the CPU 201, or a function executed by each server or each PC. Various programs to be described later are stored.

  A RAM 203 functions as a main memory, work area, and the like for the CPU 201. The CPU 201 implements various operations by loading a program or the like necessary for execution of processing from the ROM 202 or the external memory 211 into the RAM 203 and executing the loaded program.

  An input controller 205 controls input from an input device such as a keyboard (KB) 209 or a pointing device such as a mouse (not shown). A video controller 206 controls display on a display 210 such as a liquid crystal display. The display 210 is a touch panel and functions as an input device that can be operated by pressing a screen on the display.

  A memory controller 207 is provided in an external storage device (hard disk (HD)), flexible disk (FD), or PCMCIA card slot for storing a boot program, various applications, font data, user files, editing files, various data, and the like. Controls access to an external memory 211 such as a compact flash (registered trademark) memory connected via an adapter.

  A communication I / F controller 208 connects and communicates with an external device via a network, and executes communication control processing on the network. For example, communication using TCP / IP is possible.

  Note that the CPU 201 enables display on the display 210 by executing outline font rasterization processing on a display information area in the RAM 203, for example. Further, the CPU 201 enables a user instruction with a mouse cursor (not shown) on the display 210.

  Various programs to be described later for realizing the present invention are recorded in the external memory 211 and executed by the CPU 201 by being loaded into the RAM 203 as necessary. Furthermore, definition files and various information tables used when executing the program are also stored in the external memory 211, and a detailed description thereof will be described later.

  Next, a flowchart for performing control processing in the information processing apparatus 100 according to the present embodiment will be described with reference to FIG.

  FIG. 3 is a diagram illustrating an example of a flowchart for performing control processing in the information processing apparatus 100.

  The processing shown in each step shown in FIG. 3 is executed and realized by the CPU 201 of the information processing apparatus 100.

  First, when the information processing apparatus 100 starts the program in the present embodiment that can be read and executed, an inheritance setting list screen (FIG. 6) and a parameter setting screen (FIG. 7) for displaying and setting inheritance information are displayed.

  Specifically, the inheritance setting list screen (FIG. 6) includes an “inheritance destination item code” 601, “item name” 602, “inheritance source item code” 603, “item name” 604, “inheritance ratio” 605, “ It consists of items of “inheritance start” 606, “inheritance end” 607, “valid start” 608, and “valid end” 609.

The “inherited item code” 601 can accept an input from the user of the code (identification information) of the inherited item (product or the like).
The “item name” 602 can accept an input from the user of the name of the item identified by the “inherited item code” 601.

The “inheritance source item code” 603 can accept an input from the user of the code (identification information) of the item (product etc.) that is the inheritance source of the item identified by the “inheritance destination item code” 601.
The “item name” 604 can accept an input from the user of the item name identified by the “inheritance source item code” 603.

In the “inheritance ratio” 605, an input from the user of the correction magnification of the outgoing volume can be accepted. Here, the correction factor of the outgoing volume is a case where 10 collections are one item before renewal (inheritance source item), but 50 collections become one item after renewal (inheritance destination item). It can be said that 100 items (number of items) before the renewal (the number of items inherited) and 100 items (number of items) after the renewal (the number of items succeeded) are different. It can be said that the actual delivery amount of one item before renewal corresponds to 1/5 (= 10/50) of the item after renewal. For this reason, it is necessary to apply a correction factor of 1/5 to the actual delivery amount of the item before renewal. This correction magnification is referred to as the correction magnification of the outgoing volume.

  The “inheritance start” 606 can accept input from the user of date / time (time) information at the time of inheritance start indicating from which time the actual value is inherited.

  The “inheritance end” 607 can accept an input from the user of date and time (time) information at the end of inheritance indicating to what time the actual value is to be inherited.

  In this way, it is necessary to determine which period of the past issue performance (actual value) of the inheritance source item (item before renewal) is to be inherited. Because it is useless to inherit the actual issue (actual value) of the period that is not used in the prediction process, and it is useless to continue even if the issue actual amount of the inheritance source item becomes zero and no issue at all, Here, it is possible to input a setting of which period of the past issue performance (actual value) of the inheritance source item (item before renewal) to be inherited.

  The “valid start” 608 can accept input from the user of date and time (time) information at the time of valid start indicating when the record information (inherited information) set in FIG. 6 is valid.

The “valid end” 609 can accept input from the user of date and time (time) information at the time of valid end indicating how long the record information (inherited information) set in FIG. 6 is valid.
As described above, when the setting is made before the release, it is necessary to input information about when the inheritance information is valid from when.

  As described above, when the user receives input of the inheritance destination item, the inheritance source item, the inheritance period, and the expiration date of the inheritance via the screen of FIG. 6, and the save button 610 is pressed by the user. Then, the information that has received these inputs is set (step S301), and the setting contents are stored in the inheritance information DB 104.

As described above, when the inheritance destination item and the inheritance source item are set via the screen of FIG. 6, the inheritance relationship between the items is set as shown in FIG.
FIG. 5 is a diagram illustrating the inheritance relationship between items.
FIG. 5 shows the association (relationship) between the item before renewal (inheritance source item) and the item after renewal (inheritance destination item) based on the inheritance information.

  5, 501, 502, 503, 504, 505, 506, 507, and 508 are identified from the item codes of A003, A001, A006, A101, A002, S201, B191, and B199, respectively. Indicates an item.

  The inheritance relationship in FIG. 5 will be specifically described below.

  A006 (item code) is an item (inherited item) after the renewal of A003 (item code). That is, A006 is an inheritance destination item for A003, and A003 is an inheritance source item for A006.

  A101 (item code) is an item (inherited item) after renewal of A001 (item code) and A006. That is, A001 is an inheritance source item for A101.

  A201 (item code) is an inheritance item for A002 (item code), and A002 is an inheritance source item for A201. Similarly, B199 (item code) is an inheritance item for B191 (item code), and B191 is an inheritance source item for B199.

  The inheritance destination item is an item newer than the inheritance source item.

  In this manner, the inheritance relationship between the items is set on the screen of FIG.

  The parameter setting screen (FIG. 7) is a screen for setting prediction model parameters to be used for each item, and includes “item code” 701, “item name” 702, “result reference period” 703, “prediction model”. ”704,“ parameter ”705,“ prediction correction model ”706,“ correction start ”707, and“ correction end ”708.

The “item code” 701 can accept an input from a user of a code (identification information) of an item (product or the like).
The “item name” 702 can accept an input from the user of the name of the item identified by the “item code” 701.
The “result reference period” 703 can accept an input from the user of the result reference period for specifying which period of the result value is used for prediction.

The “prediction model” 704 can accept an input from a user of a prediction model (calculation formula) for calculating a prediction value, such as a moving average model, an exponential smoothing model, or a winters exponential smoothing model.

  The “parameter” 705 can accept an input from the user of the parameter of the prediction model. For example, if the prediction model is an exponential smoothing model, it is possible to accept an input from a user of a prediction model parameter such as a smoothing parameter.

  The “prediction correction model” 706 can accept an input from a user of a prediction correction model (calculation formula) for correcting a prediction value. If the predicted value is not corrected, “None” is input, and if the predicted value is corrected, the predicted correction model is input. As the prediction correction model, “upper limit model”, “linear approximation model (linear correction)”, “secondary curve model (secondary curve correction)”, “elliptical model (elliptical correction)”, and the like can be input.

  When “upper limit model” is input as the prediction correction model, it is set that the item of inheritance is an item as long as it is in stock, and “linear approximation model (linear correction)”, “quadratic curve model (quadratic curve correction)” ) ”And“ ellipse model (elliptical correction) ”, it is set that the inheritance source item is not a limited item but is an item scheduled to be discontinued. This is an application example of the setting means of the present invention.

"Start correction" 707 includes "Prediction correction model" 706, "Upper limit model", "Linear approximation model (Linear correction)", "Secondary curve model (Secondary curve correction)", "Ellipse model (Ellipse correction)" When a prediction correction model such as “)” is input, it is possible to input date and time (time) information at the start of correction indicating when the predicted value is to be corrected.

Further, in the “end of correction” 708, when a prediction correction model is input, date / time (time) information at the end of correction indicating how long the predicted value is corrected can be input.

  In the screen of FIG. 7, a prediction model can be set for each item. As another example, instead of each item, a screen that can be set for each item category (type) and prediction for each item according to the specified conditions Possible functions include setting model parameters.

  As described above, when the user accepts the input of the item, the result reference period, the prediction model, the parameter, the prediction correction model, and the correction period via the screen of FIG. 7 and the save button 709 is pressed by the user, these inputs are made. Is set (step S301), and the setting contents are stored in the prediction model DB 105.

  As described above, in step S301, the information (inheritance information) received via the screen of FIG. 6 and the information received via the screen of FIG. 7 (setting information such as prediction model parameters) are set. To do.

  Next, in step S302, the execution order of the prediction process executed for each item is determined from the inheritance relationship between the items indicated in the inheritance information.

  The order determination process in step S302 will be described in detail with reference to FIG.

  FIG. 5 is a diagram showing inheritance information, and the prediction execution order can be determined based on this inheritance information. Since the prediction execution order needs to be executed earlier as the inheritance source item, the order of the rightmost inheritance source item in FIG. 5 becomes earlier.

  It is items A001, A003, A002, and B191 that need to be predicted as the first stage.

  Since these are inheritance source items that do not become inheritance destination items, it is necessary to perform prediction at the earliest stage. The prediction execution order of these items is 1. In addition, you may perform from any item in the items of the same prediction execution order.

  The item that needs to be executed next is an inheritance-destination item for the inheritance-source item, with the item having a prediction execution order of 1 as the inheritance-source item. That is, A006, A201, and B199.

  Since the execution order of A006, which is the inheritance source of this item, has not been determined for A101, the execution order cannot be determined yet. In other words, the execution order of the items cannot be determined unless all the execution orders of the inheritance source items of the item are determined. Here, it is assumed that the prediction execution order of the items A006, A201, and B199 is 2.

  Finally, the prediction execution order of A101 is 3. Since the prediction execution order of all inheritance source items of A101 is determined, the prediction execution order of A101 can be determined. Of all inheritance source items of A101, the item with the largest execution order value is next. Since the prediction execution order of A006 is 2, the prediction execution order of A101 is 2 + 1, 3.

  Note that items that are neither the inheritance destination nor the inheritance source, that is, items that are not in the inheritance information can be executed at any time, and therefore, the prediction execution order does not matter. However, in order to distinguish from the inheritance destination item and the inheritance source item, the prediction execution order is set to 0 here.

  As described above, in step S302, the execution order of the prediction process executed for each item is determined from the inheritance relationship between the items indicated in the inheritance information.

  Next, in step S303, according to the prediction execution order determined in step S302, the information (prediction model etc.) set in step S301 and stored in the inheritance information DB 104 and the prediction model DB 105, and the results stored in the result DB 106. Using the value, the prediction process for each item is executed one by one. Detailed processing of the prediction processing in step S303 will be described later with reference to FIG.

  Next, in step S304, the prediction value obtained by the prediction process executed in step S303 is read from the prediction result DB 107 and displayed on the display 210.

  Here, an example of display on the display 210 is shown, but the predicted value may be output to be printed by a printer or the like. Step 304 is an application example of output means for outputting prediction data of the present invention.

  FIG. 10, FIG. 11, FIG. 12, FIG. 15, FIG. 16, FIG. 19, FIG. 20, and FIG. 21 are shown as examples of the predicted values displayed in step S304 and their time series graphs.

  Next, the detailed process of the prediction process of step S303 of FIG. 3 is demonstrated using FIG.

  FIG. 4 is a flowchart showing detailed processing of the prediction processing in step S303 of FIG.

  The processing of each step illustrated in FIG. 4 is executed and realized by the CPU 201 of the information processing apparatus 100.

  First, in step S401, among the items, items for which prediction processing is performed in the order determined in step S302 are set as processing targets. That is, the processing target is set in the order of items in the predicted execution order: 0, 1, 2, 3,.

  First, it is determined from the inheritance relationship (FIG. 5) specified from the inheritance information of each item whether or not the item currently being processed is an item with an inheritance source item (whether it is an inheritance destination item).

  Here, the prediction method differs depending on whether the item is an inheritance item or an item that is not, so in this step, which method is to be performed is assigned.

  And when it determines with it not being an item with an inheritance origin item (it is not an inheritance destination item) (NO), information (prediction model etc.) set by step S301 from the past track record value of the item currently processed. Thus, a future predicted value is calculated and stored in the prediction result DB 107 (step S407).

  That is, in step S407, prediction data for the inheritance source item is calculated according to the result data of the inheritance source item stored in the storage means (second calculation means).

  Detailed processing in step S407 will be described later with reference to FIGS.

  When the predicted value is calculated in step S407, the process returns to step S408.

  In step S401, if it is determined that the item currently being processed is an item that has an inheritance source item (has an inheritance source item) (is an inheritance destination item) (YES), from the inheritance relationship (FIG. 5), In step S401, the past performance values for all (one or more) inheritance source items determined to have an inheritance source item are acquired from the performance DB 106 (step S402).

  Then, the past actual value of the item currently being processed is acquired from the actual result DB 106, and for each unit period (here, each month) of the actual value, the acquired past actual value of the item currently being processed is stepped. The past actual value for the inheritance source item acquired in S402 is added to correct the past actual value for each unit period of the actual value of the item currently being processed (step S403). And the past performance value obtained by this correction | amendment is also called correction | amendment performance data.

  Also, in step S403, for each item (unit of period of actual value), using the “inheritance ratio” set in the screen of FIG. Create corrected performance data.

  For example, when the item to be processed is A201 and the inheritance source item is A002, by substituting the actual value of A201 (past performance), the actual value of A002 (past performance), and the inheritance ratio into the following formula: The correction result data of A201 (the correction result of A201) can be calculated. Further, the corrected performance data calculated here is calculated and created for each month.

  A201 correction results = A201 past results + A002 past results × succession ratio

  Here, FIG. 9 shows a graph in which the calculated corrected performance data (corrected performance) is displayed in time series for each unit period (here, each month) of the actual value.

  FIG. 9 is an example of a graph in which calculated correction performance data (correction performance) is displayed in time series for each unit period (here, each month).

  In FIG. 9, the total value of the actual value of A002, which is the inheritance source item, and the actual value of A201, which is the inheritance destination item, is plotted for each unit period (here, each month) obtained in step S403. ing.

  Next, in step S404, a prediction value for the item currently being processed is calculated using the correction result data generated in step S403 and the prediction model and prediction model parameters set in step S301 in FIG.

  That is, here, prediction data for the inheritance destination and the inheritance source item is calculated according to the combination of the performance data of the inheritance source item stored in the storage means and the performance data of the inheritance destination item (first Calculation means).

  Next, in step S405, a predicted value for the inheritance source item for the item currently being processed is acquired from the prediction result DB 107 (step S405). That is, the predicted value acquired here is the predicted data for the inheritance source item calculated according to the result data of the inheritance source item in step S407.

  In step S406, the predicted value for each month (for each unit period) predicted in step S405 is subtracted from the predicted value for each month (for each unit period) calculated in step S404 to obtain the current processing target. A predicted value (for each unit period) of each month for the item (inherited item) is calculated and stored in the prediction result DB 107.

  For example, when the current processing target item is A201, by substituting the correction result data of A201 (correction result of A201), the predicted value (prediction result) of A002 that is the inheritance source item, and the inheritance ratio into the following equation: , A201 prediction result (predicted value) can be calculated. Further, the predicted value (also referred to as predicted data) calculated here is calculated and generated for each month.

  A201 prediction result = A201 prediction result calculated from the correction result of A201−A002 prediction result × inheritance ratio

  Next, in step S408, it is determined whether there is a next item according to the order determined in step S302. If it is determined that there is a next item (YES), the next item is processed. Then, the process returns to step S401. On the other hand, if it is determined that there is no next item, the prediction process is terminated, and the process proceeds to step S304.

  Next, a specific example of the processing of FIG. 4 will be described below.

  First, with respect to an item not in the inheritance information for which the prediction execution order is 0, since the item has no inheritance source in step S401, the process proceeds to step S407.

  In S407, a prediction result is calculated using the prediction model and its parameters specified with reference to the past shipping results of this item. In S408, if there is a next item, the process returns to S401, and the next item is predicted.

  Next, processing of an item whose prediction execution order is 1 will be described. Since the item whose prediction execution order is 1 is not the inheritance destination but the item that is the inheritance source, the prediction is performed in step S407, and the same processing as the item whose execution order is 0 is performed.

  Next, processing of items whose prediction execution order is two or more will be described. Since these items have inheritance source items, it is determined in S401 that the process proceeds to S402.

  Regarding the processing after S402, items A002 (505) and A201 (506) in FIG. 5 will be described as an example. Since the prediction execution order of A002 is 1, as described above, the prediction process has already been executed in step S407. Since A201 has a prediction execution order of 2, the description will proceed centering on A201.

  FIG. 8 is an example of monthly sales data (time series data obtained by totaling actual values in time series) of the inheritance source item A002 and the succession destination item A201 for 36 months from January 2008 to December 2010. is there.

  FIG. 9 is a graph of the time series data of FIG.

  The inheritance source item A002, the inheritance destination item A201, and their total value are displayed in a graph.

  It can be seen that the delivery record amount of the inheritance destination item A201 has started to be issued from November 2010, and that the inheritance source item A002 has decreased the delivery record amount from November 2010. In addition, although it is clear at a glance when seen in the graph, it can be seen that the total amount of issued goods of the inheritance source item A002 and the inheritance destination item A201 has a tendency.

  Next, in step S402, the past result data of A002 as the inheritance source is acquired. In step S403, the past record of A002 is added to the past record of A201 as the inheritance destination to create a corrected record of A201. At this time, using the inheritance ratio set on the inheritance setting screen, the correction result of A201 is created for each month by the following formula. In addition, the correction performance of A201 is the total graph of FIG.

  A201 correction results = A201 past results + A002 past results × succession ratio

  Next, in step S404, prediction is executed using the specified prediction model and its parameters using the correction results created in step S403.

  FIG. 10 is an example of a graph in which the corrected actual value of the inherited item A201 and the prediction result using the corrected actual value are represented in time series.

  Next, in step S405, the already executed prediction result of the inheritance source A002 is acquired. FIG. 11 is a graph of the prediction result of the inheritance source item A002.

  In step S406, the prediction result of the inheritance source item A002 acquired in S405 is subtracted from the prediction result calculated in S404. The calculation formula is as follows.

  A201 prediction result = A201 prediction result calculated from the correction result of A201−A002 prediction result × inheritance ratio

  FIG. 12 is an example of a graph in which the past results (sales actual value) of A201, which is the inheritance destination item, and the prediction results are represented in time series.

  In this way, the items whose prediction execution order is two or more are predicted according to the steps of S402 to S406, the process proceeds to the prediction process for the next item, and after the prediction of the last item is completed, all items are predicted in S408. When the prediction execution is completed, the process ends. As described above, in the prediction process, in order to refer to the prediction result of the inheritance source item in step S405, it is necessary to perform prediction from the inheritance source item.

  Next, detailed processing in step S407 in FIG. 4 will be described with reference to FIG.

  FIG. 13 is an example of a flowchart of detailed processing in step S407 of FIG.

  In step S301 of FIG. 3, “None”, “Upper limit model”, “Linear approximation model (Linear correction)”, “Secondary curve model (Secondary curve correction) are selected as the prediction correction model 706 via the screen of FIG. ) ”Or“ ellipse model (elliptical correction) ”is input, and a prediction correction model 706 is set for each item.

  In step S407, it is determined which prediction correction model is currently set for the item to be processed, and the prediction processing method executed in step S407 is switched and executed according to the determination result.

  FIG. 13 shows “prediction model (straight line correction)”, “second-order curve model (second-order curve correction)”, and “elliptical model (elliptical correction)” as prediction correction models 706 via the screen of FIG. It is a flowchart of the process performed when it determines with either being set.

  In other words, the prediction process shown in FIG. 13 is executed as a prediction process for an inheritance source item that has been determined to end sales.

  If it is determined that “None” is set as the prediction correction model 706 via the screen of FIG. 7, the processing of step S1301 is not performed without performing the processing of steps S1302 to S1306 described later. Execute.

  The process shown in each step shown in FIG. 13 is executed and realized by the CPU 201 of the information processing apparatus 100.

  First, in step S1301, using the prediction model set in step S301 and the actual value for the item acquired from the actual DB 106 for the currently processed item, the predicted value is obtained via the screen of FIG. Is calculated.

  Next, in step S1302, the correction rate at each time point from the correction start point 707 to the correction end point 708 set in FIG. 7 is calculated from the prediction correction model 706 set via the screen in FIG. The correction factor calculated here is stored in a storage unit such as the external memory 211. The calculated correction factor is an application example of the correction information of the present invention.

  In step S1303, the correction start time 707 is set as a processing target (0 is substituted into the variable i).

  Next, in step S1304, a corrected predicted value is calculated by multiplying the predicted value calculated in step S1301, which is the current processing target, by the correction rate calculated in step S1302, and this is calculated as the current processing target. Use the predicted value at the time. Here, each time point represents each unit period such as each month, for example.

  That is, in step S1304, a process of correcting the prediction value calculated in step S1301 that is currently processed by the correction factor is performed. Step S1304 is an application example of the correcting means of the present invention.

  In step S1305, it is determined whether or not the prediction values calculated in step S1301 for all time points from the correction start time point 707 to the correction end time point 708 set in FIG. When the predicted values for all the time points are not corrected (NO), the time point to be processed is set to the next time point (adding 1 to i) (step 1306), and the process of step S1304 is executed. On the other hand, when the prediction values for all the points up to the correction end point 708 are corrected (YES), the process of FIG. 11 is ended.

  A specific example of the flowchart shown in FIG. 11 will be described below.

  This will be described using the data of the inheritance source item A002 in FIG.

  The correction start time is January 2011, and the correction end time is May 2011.

  In step S301, a result reference period, a prediction model, a parameter, and the like are set for each item via FIG. Here, in order to perform the processing of FIG. 11, parameters such as a prediction correction model, correction start, and correction end are set.

  As described above, for example, a straight line (straight line correction), a quadratic curve (quadratic curve correction), an elliptic curve (elliptic correction), or the like as shown in FIG. 14 can be considered as a correction factor calculation method for correcting the predicted value. .

FIG. 14 is a graph showing the correction rate at each point in time from the correction start point (November 2010) to the correction end point (May 2012), obtained from each prediction correction model (calculation formula). It is an example.
As a specific example, straight line correction is used here.

  FIG. 15 shows, in time series, the actual value of the inheritance source item A002, the predicted value calculated from the actual value, and the predicted value (corrected predicted value) obtained by correcting the predicted value using the prediction correction model. It is an example of the represented graph.

  In FIG. 15, the values plotted as predictions are the prediction values calculated in S1301.

  Then, since the prediction correction method is a straight line, the correction start time is January 2011, and the correction end time is May 2011, the correction rate is calculated in S1302.

  The correction rate is the Y-axis value shown in FIG.

  In S1303 to S1305, the predicted value for each month is corrected.

  For example, when 360, which is the predicted value for February 2011, is multiplied by the correction rate of 0.5, 180 is obtained as the corrected predicted value. In this way, the corrected predicted value corrected by multiplying the predicted correction rate for each month by the predicted value corresponds to the corrected prediction in FIG. It can be seen that the corrected predicted value gradually decreases so that the predicted value in May 2011 becomes zero.

  In this way, it is possible to calculate the predicted value for the inheritance source item that has been determined to end selling.

  Then, the process from step S402 to step S406 shown in FIG. 4 is performed for the inheritance destination item, and the prediction value for the inheritance destination item is calculated. That is, from the prediction value calculated in step S404 in step S406, FIG. By executing this process, the predicted value of the inheritance destination item is calculated by subtracting the predicted value of the inheritance source item that has been determined to end the sale.

  Thus, by calculating the predicted value, it is possible to further increase the prediction accuracy.

  For example, the prediction of the inheritance source item A201 is performed by performing the prediction of the inheritance source item A002 as shown in FIG. 15, and subtracting the prediction result of the inheritance source item A002 from the prediction result of the inheritance destination and the inheritance source in FIG. Corresponds to the graph of FIG. It can be seen that the accuracy is further improved compared with the prediction result of FIG.

  FIG. 16 is an example of a graph showing the predicted values of the inheritance item A201 in time series.

  Next, detailed processing in step S407 in FIG. 4 will be described with reference to FIG.

  FIG. 17 is an example of a flowchart of detailed processing in step S407 of FIG.

  In step S301 of FIG. 3, “None”, “Upper limit model”, “Linear approximation model (Linear correction)”, “Secondary curve model (Secondary curve correction) are selected as the prediction correction model 706 via the screen of FIG. ) ”Or“ ellipse model (elliptical correction) ”is input, and a prediction correction model 706 is set for each item.

  In step S407, it is determined which prediction correction model is currently set for the item to be processed, and the prediction processing method executed in step S407 is switched and executed according to the determination result.

  FIG. 17 is a flowchart of processing executed when it is determined that the “upper limit model” is set as the prediction correction model 706 via the screen of FIG. 7.

  That is, the prediction process shown in FIG. 13 is executed as a prediction process for an inheritance source item whose sale has been determined to end, and which will be terminated when the current inventory is sold.

  The process shown in each step shown in FIG. 17 is executed and realized by the CPU 201 of the information processing apparatus 100.

  First, in step S1701, using the prediction model set in step S301 and the actual value for the item acquired from the actual DB 106 for the currently processed item, the predicted value is obtained via the screen of FIG. Is calculated.

  In step S1702, the prediction start point (prediction point) is set as a processing target.

  In step S1703, a prediction upper limit value at the prediction point of the current processing target is calculated. Specifically, the value obtained by adding the number of stocks scheduled to be received to the current processing target prediction time to the number of stocks immediately before the current processing target prediction time is used as the prediction at the current processing target prediction time. Calculate as the upper limit.

  That is, a predicted upper limit value indicating the number of stocks that can be issued is calculated according to the number of stocks scheduled to be stored at the predicted time point and the stock number at the previous predicted time point (upper limit value calculating means).

  Next, it is determined whether or not the predicted value calculated in step S1701 at the current processing target time is larger than the upper limit predicted value calculated in step S1703 (step S1704). If it is determined that the predicted value calculated in step S1701 is larger than the upper limit predicted value calculated in step S1703 (YES), the upper limit predicted value calculated in step S1703 is calculated in step S1701. It is set as the predicted value (step S1705). On the other hand, if it is determined that the predicted value calculated in step S1701 at the time of the current processing target is not more than the upper limit predicted value calculated in step S1703 (NO), the process proceeds to step S1706.

  In step S1704, it is determined whether the prediction upper limit value calculated in step S1703 exceeds the prediction data calculated in step S1701 (determination means).

  When the process of step S1705 is executed, the process proceeds to step S1706.

  In step S1706, it is determined whether or not the processing from step S1703 to step S1705 has been performed for all time points from the prediction start time point to the prediction end time point, and processing is performed for all time points from the prediction start time point to the prediction end time point. When it is determined that the process has been performed (YES), the process of FIG. 17 is terminated. When it is determined that there is a time point when the process has not yet been performed (NO), the processing target is set to the next time point (i. 1 is added (step S1701), and the process returns to step S1703.

  Next, a specific example of the flowchart shown in FIG. 17 will be described.

  This will be described using the data of the inheritance source item A002 in FIG.

  In step S1701, as described above, prediction is performed using the prediction model set in step S301 and the actual value for the item acquired from the actual DB 106 for the item currently processed (for example, the inheritance source item A002). Calculate the value.

  FIG. 18 shows the performance (actual value) in December 2010, the forecast (predicted value) after January 2011, the stock (number of stock) at the end of December 2010, and 2011 It is an example of the data table which represented the warehousing schedule (stock warehousing value) after January in time series.

  The predicted values (predicted values from January to June 2011) calculated at step S1701 are stored in the data table of FIG.

  First, in step S1702, the processing target is set to the time of January 2011, and the processing is started from this time.

  Next, in step S1703, a predicted upper limit value for January 2011 is calculated.

  The predicted upper limit value is the stock amount at the end of the month when there is no sale in the current month to be processed. In other words, the predicted upper limit value is a value obtained by adding the stock amount (stock) at the end of last month to the stock amount (scheduled stock) of the current month. Therefore, a value (350) obtained by adding the stock quantity 200 at the end of December 2010 to the planned number of goods received 150 in January 2011 becomes the predicted upper limit value in January 2011.

  In the determination process in step S1704, the predicted value for January 2011 is 310, and this value is determined to be a value equal to or lower than the predicted upper limit (350). In step S1705, the predicted value for January 2011 is determined. 310 is a corrected predicted value (corrected prediction).

  The data table in which the prediction upper limit value calculated in this way and the corrected prediction value are additionally registered in FIG. 18 is the table in FIG.

FIG. 19 shows the actual value of item A002 as the inheritance source in December 2010, the predicted value after January 2011, the number of stocks (actual value) at the end of December 2010, and the scheduled arrival of goods after January 2011 It is an example of the data table which represented the value, the prediction upper limit value after January, 2011, and the correction | amendment prediction value after January, 2011 yen in time series.
Return to explanation.

  Next, it is determined in step 135 that the process has not been performed until the prediction end point (June 2011), and the next month, February 2011, is set as the process target (step S1707), and the process returns to step S1703.

  Next, in step S1703, a predicted upper limit value for February 2011 is calculated.

  Specifically, by calculating “stock quantity at end-December 2010 (200) + estimated warehousing quantity in January 2011 (150) −corrected forecast value in January 2011 (310)”, the end of January 2011 Is calculated to be 40.

  Then, similarly to the above-described processing, by calculating “stock quantity at the end of January 2011 (40) + stock receipt quantity in February 2011 (100)”, the forecast upper limit value (140) for February 2011 is calculated. can get.

  Next, in step S1704, it is determined that the predicted value (363) in February 2011 is larger than the predicted upper limit value (140) in February 2011. In step S1705, the predicted correction value in February 2011 is The prediction upper limit value is 140.

  Such processing is performed for each month, and the predicted value for each month is corrected. The result is the value of the corrected prediction row in FIG. FIG. 20 is a graph showing the corrected predicted value.

  Here, the predicted value of the inheritance source item A002 calculated in step S1701 is used in the subsequent processing as the predicted prediction value corrected in step S1704 and step S1705 as the predicted value of the inheritance source item A002. Become.

  In this way, it is possible to calculate a predicted value for an inheritance source item that will be discontinued if the current inventory is sold.

  And the process of step S402 to step S406 shown in FIG. 4 is performed about an inheritance destination item, and the predicted value about an inheritance destination item is calculated. That is, in step S406, the prediction value for the inheritance source item that will be sold out when the current inventory is sold is obtained by subtracting the prediction value calculated in step S404 from the prediction value in step S404, Calculate the predicted value of the inheritance item.

  Thus, by calculating the predicted value for the inheritance item, it is possible to further increase the prediction accuracy.

  Prediction at each time point in the graph of FIG. 10 in which the predicted value of the total of the inheritance destination item and the inheritance source item obtained by performing prediction using the total value of the actual value of the inheritance destination item and the actual value of the inheritance source item is plotted. The value obtained by subtracting the corrected predicted value at each time point of the inheritance source item A002 in FIG. 20 from the value becomes the predicted value (prediction result) at each time point of the inheritance destination item A201.

  FIG. 21 is a graph in which predicted values of the inheritance destination item A201 obtained as a result of the prediction of the inheritance destination item A201 are plotted in time series.

  In other words, the inheritance destination item and the inheritance source item as a whole are sold as shown in FIG. 10, and the inheritance source item is sold out as much as it is in stock, and the remainder is sold as a new product after renewal. Is shown.

  As described above, according to the present invention, the prediction data of the inheritance destination item is obtained according to the prediction result obtained according to the performance data of the inheritance source item and the succession destination item, and the prediction result of the inheritance source item obtained according to the performance data of the inheritance source item. By calculating, the prediction accuracy of the prediction data of the inheritance item can be improved.

  Further, according to the present invention, the prediction accuracy of the product after the renewal can be improved by referring to the prediction result of the product before the renewal based on the inheritance information of the product before the renewal and the product after the renewal.

  As mentioned above, the past performance of the product before and after the renewal is used, and the past information of the product before the renewal is incorporated into the past performance of the product after the renewal using the association information (inherited information). Information can be supplemented. In the prediction of the product after the renewal, it is possible to carry out the overall prediction before and after the renewal by making a prediction based on the total results before and after the renewal.

  Further, by subtracting the prediction result before the renewal from the prediction result, the prediction result of the product after the renewal can be derived.

  However, since it is necessary to determine the prediction result before the renewal before performing the prediction of the product after the renewal, the order of performing the prediction is determined in advance.

  In addition, since it is possible to accurately predict the end-of-sale trend of the product before the renewal (the sales gradually decline), it is possible to accurately predict the sales of the product after the renewal.

  Another object of the present invention is to supply a storage medium storing software program codes for realizing the functions of the above-described embodiments to a system or apparatus, and the computer (or CPU or MPU) of the system or apparatus stores the storage medium. Needless to say, this can also be achieved by reading and executing the program code stored in.

  In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiments, and the program code itself and the storage medium storing the program code constitute the present invention.

  As a storage medium for supplying the program code, for example, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a CD-R, a magnetic tape, a nonvolatile card, a ROM, or the like can be used.

  Further, by executing the program code read by the computer, not only the functions of the above-described embodiments are realized, but also an OS (basic system or operating system) running on the computer based on the instruction of the program code. Needless to say, a case where the function of the above-described embodiment is realized by performing part or all of the actual processing and the processing is included. An information processing apparatus (computer) that can read and execute the program code realizes the functions of the above-described embodiments.

  Further, after the program code read from the storage medium is written in a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, the function is determined based on the instruction of the program code. It goes without saying that the CPU or the like provided in the expansion board or function expansion unit performs part or all of the actual processing and the functions of the above-described embodiments are realized by the processing.

DESCRIPTION OF SYMBOLS 100 Information processing apparatus 101 Setting means 102 Prediction means 103 Display means 104 Inheritance information DB
105 Prediction model DB
106 Achievement DB
107 Prediction result DB
108 Order determining means 211 External memory

Claims (6)

Storage means for storing the result data of the inheritance source item and the result data of the inheritance destination item;
A first calculation for calculating prediction data for the inheritance destination and the inheritance source item according to a combination of the performance data of the inheritance source item and the performance data of the inheritance destination item stored in the storage unit Means,
Second calculation means for calculating prediction data for the inheritance source item according to the result data of the inheritance source item stored in the storage means;
Third calculation means for calculating prediction data for the inherited item by subtracting the prediction data calculated by the second calculation means from the prediction data calculated by the first calculation means;
Output means for outputting the prediction data calculated by the third calculation means;
An information processing apparatus comprising: The storage means further stores correction information for correcting the prediction data,
A correction unit for correcting the prediction data for the inheritance source item calculated by the second calculation unit according to the correction information stored in the storage unit;
The third calculation means subtracts prediction data for the inheritance source item corrected by the correction means from the prediction data calculated by the first calculation means, thereby obtaining the inheritance destination item. The information processing apparatus according to claim 1, wherein prediction data is calculated. An upper limit calculating means for calculating a predicted upper limit value indicating the number of stocks that can be issued according to the number of stocks scheduled to be received at the forecast time and the stock numbers at the previous forecast time;
A determination unit that determines whether the prediction upper limit value calculated by the upper limit value calculation unit exceeds the prediction data calculated by the second calculation unit;
The third calculating means, when the determining means determines that the predicted upper limit value calculated by the upper limit value calculating means exceeds the predicted data calculated by the second calculating means, 2. The prediction data for the inherited item is calculated by subtracting a prediction upper limit value determined to exceed the prediction data from the prediction data calculated by one calculation means.
Or the information processing apparatus of 2. Further comprising setting means for setting whether the inheritance source item is an item that is limited to stock or an item that is not a limited item but is scheduled to be discontinued;
When the setting means sets that the item to be inherited is an item as long as it is in stock, the third calculating means determines that the predicted upper limit value calculated by the upper limit value calculating means is determined by the determining means. When it is determined that the prediction data calculated by the second calculation means is exceeded, the prediction upper limit value determined to exceed the prediction data is subtracted from the prediction data calculated by the first calculation means. When the prediction data for the inheritance destination item is calculated, on the other hand, the setting means sets that the inheritance source item is not a limited item but is scheduled to be discontinued. The third calculating means predicts the inheritance source item corrected by the correcting means from the prediction data calculated by the first calculating means. By then pulling the information processing apparatus according to any one of claims 1 to 3, characterized in that to calculate the prediction data for said inheritance destination item. A control method for an information processing apparatus comprising storage means for storing result data of an inheritance source item and result data of an inheritance destination item,
The first calculation means of the information processing apparatus, according to a combination of the result data of the inheritance source item stored in the storage means and the result data of the inheritance destination item, the inheritance destination and the inheritance source A first calculation step of calculating prediction data about the item;
A second calculation step in which a second calculation unit of the information processing apparatus calculates prediction data for the inheritance source item according to the result data of the inheritance source item stored in the storage unit;
The third calculation means of the information processing apparatus subtracts the prediction data calculated in the second calculation step from the prediction data calculated in the first calculation step, thereby A third calculation step for calculating prediction data;
An output step in which the output means of the information processing apparatus outputs the prediction data calculated in the third calculation step;
A control method comprising: It is a program that can be read and executed by an information processing apparatus that includes storage means for storing the result data of the inheritance source item and the result data of the inheritance destination item,
The information processing apparatus;
A first calculation for calculating prediction data for the inheritance destination and the inheritance source item according to a combination of the performance data of the inheritance source item and the performance data of the inheritance destination item stored in the storage unit Means,
Second calculation means for calculating prediction data for the inheritance source item according to the result data of the inheritance source item stored in the storage means;
Third calculation means for calculating prediction data for the inherited item by subtracting the prediction data calculated by the second calculation means from the prediction data calculated by the first calculation means;
A program that functions as output means for outputting prediction data calculated by the third calculation means.

Images