JP5079901B1 - Prediction method and prediction device for changes in work volume and number of visitors - Google Patents

Abstract

An object of the present invention is to calculate a predicted value of the amount of work and the number of visitors for the next month.
SOLUTION: A step of extracting data relating to office work volume personnel, monthly office work volume, and peak index from N months to M months before the forecast target month, and data relating to the extracted office work volume personnel and monthly office work volume Calculating an average value of each item, a step of calculating an increase / decrease rate of each item, a step of multiplying the calculated increase / decrease rate by a value K months before the prediction target month, and N month of the prediction target month Assigning code values according to the characteristics of the day to each date from the previous month to M months ago and each date of the prediction target month, and code values from N months to M months before the assigned prediction target month Determining whether there is a matching code value, extracting the peak index for each item for all dates that match the determined code value, and calculating the average value of the peak index for each extracted item Do And a step. However, N, M, and K are natural numbers that satisfy N>K> M.
[Selection] Figure 8

Description

  The present invention relates to a prediction method and a prediction device for a change in the amount of work, a change in the number of visitors, and more specifically, in a computer connected to a database, the amount of work in the next month, a change in the amount of visit to predict the number of visitors, and a change in the number of visitors The present invention relates to a prediction method and a prediction apparatus.

  Conventionally, the daily work volume and the number of visitors for the next month are forecasted based on past performance data and the experience of managers, etc., and the waiting time forecast for each day and time zone is based on whether the store is in the city center. It is done using a congestion inquiry system that outputs daily congestion forecast information as a busy, busy, busy pattern that divides stores by type depending on whether it is in a residential area, etc. ing.

  In addition, as a system for predicting the number of customers at retail stores, restaurants, etc., daily report information, labor information, sales information, weather sensor information, and network information are input, a Bayesian network probability table is created, and the number of visitors is predicted. A visitor number prediction system using a Bayesian network that outputs data is known (see, for example, Patent Document 1).

  Furthermore, as a system for providing visitor prediction / actual data, when a destination is registered in the navigation system, route guidance information, location information, and route guidance start date / time are accumulated, and when the route guidance by the navigation system ends, the destination Arrival prediction information, route guidance end date and time are accumulated, and the stored route guidance information, location information, route guidance start date and time, destination arrival presence / absence information, route guidance end date and time are stored as visitor prediction / result information providing system (See, for example, Patent Document 2).

Japanese Patent Laying-Open No. 2005-228014 (page 5-7, FIG. 1) JP 2010-191763 (page 4-6, FIG. 2)

  However, if you divide stores by type and output daily congestion forecast information as a busy pattern patterned by segmented type, it is difficult to create all patterns that can be considered for effective forecasting There was a problem of being.

  If it is difficult to apply the busy pattern to output the daily congestion forecast information as a busy pattern that is classified by type and classified into stores, it is necessary to manually correct the daily congestion forecast information There was also a problem that there was.

  In addition, if you divide stores by type and output daily congestion forecast information as a busy pattern that is patterned by sorted type, you can only output busy, flat, busy busy patterns as forecast results, There was also a problem that the degree of busyness and busyness could not be output.

  The present invention has been made in view of such problems. The purpose of the present invention is to calculate the amount of work for the next month, the predicted value of the number of customers, It is to provide a prediction device.

  In order to achieve such an object, the present invention provides a database in which data relating to past work volume personnel, monthly work volume, and peak index are stored monthly. On the connected computer, data on the amount of work from N to M months before the forecast month from the database, data on monthly work and peak index, where N and M are natural numbers and N> (M + 1) The step of calculating the data, the step of calculating the average value of each item of the extracted work volume personnel and the monthly work volume, and M months before the prediction target month extracted by the calculated average value By calculating the rate of increase / decrease for each item by dividing the value of, and for each item of data on the extracted work volume personnel and monthly work volume, the calculated increase / decrease rate was extracted A value of K months before the measurement month, where K is a natural number and satisfies N> K> M, and the multiplied value is stored as a predicted value for the amount of work and the amount of monthly work; A step of assigning code values according to the characteristics of the day to each date from N months to M months before the prediction target month and each date of the prediction target month, and a code value of each date of the assigned prediction target month Determining whether there is a code value that matches the code values from N months to M months before the assigned prediction target month, and from N months before the extracted prediction target month to M months The step of extracting the peak index of each item for all dates that match the determined code value from the previous peak index data, and calculating the average value of the peak index of each extracted item Average value for peak index And storing as a prediction value, characterized by comprising a predicted value related to the stored office amount personnel and monthly office amount, and outputting a predicted value related to the stored peak index.

  According to this configuration, it is possible to calculate the amount of work for the next month and the predicted value of the number of visitors, and provide the prediction result of the amount of work and the number of visitors.

  The invention described in claim 2 is characterized in that the predicted value related to the stored peak index according to claim 1 includes a predicted value related to the peak index related to the number of visitors.

  According to this method, it is possible to calculate predicted values for the amount of work and the number of visitors for the next month, and to provide prediction results of the amount of work and the number of visitors.

  The invention according to claim 3 is a computer connected to a database that stores data on past work volume personnel, monthly work volume, and peak index on a monthly basis. Calculates the average value of each item of data on the amount of work volume from the previous to M months ago, the means for extracting the data on the monthly work volume and the peak index, and the data on the extracted work volume personnel and the monthly work volume Means, means for calculating an increase / decrease rate of each item by dividing the value of M months before the target month extracted by the calculated average value, and the extracted amount of work and monthly amount of work Means for multiplying the calculated increase / decrease rate for each item of data by the value K months before the extracted prediction target month, and storing the multiplied value as a predicted value relating to the amount of work and the monthly amount of work When, Means for allocating code values according to the characteristics of the day to each date from N months before M month to M month and each date of the prediction target month, and code value of each date of the allocated prediction target month , Means for determining whether there is a code value that matches the code values from N months before M months before the allocated prediction target month, and M months from N months before the extracted prediction target month Calculated by calculating the average value of the peak index of each extracted item and the means to extract the peak index of each item for all dates that match the determined code value from the previous peak index data Means for storing an average value as a predicted value for a peak index, means for outputting a predicted value for stored work volume personnel and monthly work volume, and a predicted value for stored peak index And

  According to this configuration, it is possible to calculate the amount of work for the next month and the predicted value of the number of visitors, and provide the prediction result of the amount of work and the number of visitors.

  The invention described in claim 4 is characterized in that the predicted value related to the stored peak index according to claim 3 includes a predicted value related to the peak index related to the number of visitors.

  According to this apparatus, it is possible to calculate a predicted value of the next month's office work volume and the number of visitors, and provide a prediction result of the office work volume transition and the visitor number transition.

  As described above, according to the present invention, the step of extracting data related to the work amount personnel, the monthly work amount, and the peak index from N months to M months before the target month is extracted from the database. Calculating the average value of each item of data regarding the amount of work and the amount of monthly work, and dividing the value of each item by M times of the forecast target month extracted by the calculated average value For each item of the data relating to the step of calculating the rate of increase / decrease and the extracted work volume personnel and monthly work volume, the calculated rate of increase / decrease is multiplied by the value K months before the extracted forecast target month, Steps for storing the multiplied value as a predicted value related to the amount of work and the amount of work for the month, each date from N months to M months before the prediction target month, and each date of the prediction target month Assign code values according to And a step of determining whether or not there is a code value that matches a code value from N months to M months before the assigned prediction target month, for the code value of each date of the assigned prediction target month And extracting the peak index of each item of all the dates with the determined code values from the data relating to the peak index from N months to M months before the extracted forecast target month, The average value of the peak index for each item is calculated, the calculated average value is stored as a predicted value for the peak index, the predicted value for the stored work volume personnel and the monthly work volume, and the stored peak index And a predicted value of the number of visitors can be calculated in the next month.

It is a block diagram which shows the system environment which estimates the office work volume transition and visitor number transition concerning one Embodiment of this invention. It is a block diagram which shows the prediction apparatus of the business amount transition concerning one Embodiment of this invention, and a visitor number transition. It is a flowchart which shows the prediction method of the office work volume transition and visitor number transition concerning one Embodiment of this invention. It is a flowchart which shows the prediction method of the office work volume transition and visitor number transition concerning one Embodiment of this invention. It is a flowchart which shows the prediction method of the office work volume transition and visitor number transition concerning one Embodiment of this invention. It is a figure which illustrates the group and code value which are used for the prediction method of the office work volume transition and the visitor number transition concerning one Embodiment of this invention. It is a figure which illustrates the example to which the group and code value which are used for the prediction method of the amount-of-work volume transition and the visitor number transition concerning one Embodiment of this invention were applied. It is a figure which illustrates the prediction result of the amount-of-work transition and visitor number transition concerning one Embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a configuration diagram showing a system environment for predicting a change in the amount of work and a change in the number of visitors according to an embodiment of the present invention. The prediction device 101 that calculates and outputs the prediction result of the business amount transition and the visitor number transition, and the database 102 that stores the data for predicting the business amount transition and the visitor number transition are connected by a connection cable. .

  In this specification, for the purpose of explanation, the system environment directly connects the prediction apparatus 101 and the database 102, but the prediction apparatus 101 and the database 102 may be connected via the network 103. Further, the prediction apparatus 101 may be accessed from another computer 104 via the network 103 to calculate and output the prediction result. A prediction result may be calculated and output by accessing the prediction apparatus 101 from another computer 106 via the Internet 105. Further, in the present specification, for the purpose of explanation, the prediction environment 101 is separated from the database 102 in the system environment, but the prediction apparatus 101 may include the database 102.

  The prediction device 101 processes an application described later with reference to FIG. 2 to connect to the database 102 and search for data stored in the database 102 to predict a change in the amount of work and a change in the number of visitors. Data necessary for the extraction is extracted, and a predicted value is calculated from the extracted data. Details will be described later with reference to FIGS.

  The database 102 stores, for example, a business efficiency management table, a business efficiency management schedule, and a peak day summary table as form data for recording past performance values for each month. The administrative efficiency management table includes, for example, a high counter-related office counter, the back, and the chief administrative staff, and a low counter-related office consultation counter, various notification counters, the rear, and the chief administrative staff. However, the work volume personnel is a numerical value obtained by converting the standard processing time when performing paperwork according to the paperwork procedure, and the unit is the number of people.

  The work efficiency management schedule includes, for example, deposits, withdrawals, counter telegraph transfers, counter document transfers, tax payments, yen exchange, bankbook entry / carry-back work volume for high counter related office work, and low counter related office work. New account savings account, normal account cancellation, periodic new account, periodic continuation, periodic cancellation, passbook entry / carry-over, account transfer acceptance, loss acceptance / loss completion, address change / name change, foreign currency sale, foreign currency purchase, T / C Includes sales and T / C purchases. However, the amount of work is the number of cases per business day.

  The peak day summary table includes items for each day, for example, high-counter related deposits, withdrawals, wire transfers, document transfers, tax payments, yen exchanges, passbook entry / carry forward work volume peak indices, and low counter-related items. New deposits, regular account cancellations, regular new account / continuation, periodic cancellation, receipt of notifications, foreign currency exchange, peak index of T / C work volume and high / low (consultation) at the lobby corner window , Including the peak index of the number of visitors in the low (notifications), the peak index of the number of visitors in the automatic service machine in the lobby corner, and the number of counters in the high, low (consultation) and low (various notifications).

However, the peak index is a value obtained by dividing the current day value by the average value during the month and multiplying the divided value by 100. In other words, the peak index is an index representing busyness with the average of one business day per month as 100. is there. For example, assuming that the average number of visitors per business day per month is 50 and the number of visitors on the day is 75, the peak index of the number of visitors on the day is as follows.
Peak index = 75 people / 50 people x 100 = 150
FIG. 2 is a block diagram showing an apparatus for predicting a change in the amount of work and a change in the number of visitors according to an embodiment of the present invention. The prediction device 101 includes a central processing unit 201 that controls each block, a memory 202 that stores programs and data, a storage device 203 that stores data, an input device 204 that processes input, and a screen display that displays output. A device 205 and a network connection device 206 connected to the database 102 are connected to each other via a bus.

  The central processing unit 201 executes a program stored in the memory 202. When the central processing unit 201 receives a request for prediction of the transition of the business amount and the transition of the number of visitors from the input device 204, the central processing unit 201 connects to the database 102 via the network connection device 206 by the basic software 211, the application 212, and the data 213. Then, the data in the database 102 is searched to extract data necessary for predicting the business amount transition and the visitor number transition, the predicted value is calculated from the extracted data, and the calculated predicted value is stored in the screen display device 205. Display.

  With such a configuration, the predicted value of the next month's amount of work and the number of visitors is calculated, and the prediction result of the change in the amount of work and the number of visitors is provided.

  FIG. 3 is a flowchart showing a method for predicting the transition of the amount of work and the transition of the number of visitors according to the embodiment of the present invention. Processing of the prediction method in the case where the application 212 stored in the memory 202 of the prediction device 101 is executed to acquire the prediction result of the business amount transition and the visitor number transition will be described.

  When the application 212 stored in the memory 202 of the prediction device 101 is executed and a request for prediction of business amount transition and visitor number transition is received from the input device 204 of the prediction device 101, the business amount is received from the database 102 in step S301. The form data necessary for predicting the transition and the number of visitors is extracted, and the process proceeds to step S302. For example, the form data extracted in step S301 includes a business efficiency management table, a business efficiency management schedule, and a peak day overview table.

  In step S <b> 302, it is determined whether or not all the form data for (N−M + 1) months from N months to M months before the prediction target month have been extracted from the database 102. Note that N and M are natural numbers and satisfy N> (M + 1). For example, if N is 15, M is 3, and the prediction target month is December 2010, all the form data for 13 months from September 2009 to September 2010 is extracted from the database 102. Determine if you did.

  In step S302, when the form data for (N−M + 1) months from N months before M months before the prediction target month has been extracted from the database 102, the process proceeds to step S303. In step S303, the average value from N months before the prediction target month to M months before is calculated for items whose number of staff to be placed is proportional to the amount of paper among the items of the paperwork details as the amount of paperwork.

  For example, in step S303, the items in which the number of staff to be allocated is proportional to the amount of work include the counter for high counter-related work, the back, and the chief work amount of staff, and the counter for clerical work related to the low counter, various notification windows, the back, and the chief The amount of work in the office, and deposits, withdrawals, bank wire transfers, bank document transfers, tax payments, yen exchange, bankbook entry / carry-over work volumes, and low counter-related work Termination of ordinary deposit, periodic new account, periodic continuation, periodic termination, bankbook entry / carry-over, account transfer acceptance, loss acceptance / loss completion, address change / name change, foreign currency sale, foreign currency purchase, T / C sale, T / C purchase Can be included.

  Further, in step S303, an average value from N months to M months before the prediction target month is calculated for the monthly work amount item, and the process proceeds to step S304. For example, monthly work volume items include high-counter related deposits, withdrawals, wire transfers, document transfers, tax payments, yen exchanges, passbook entry / carry-over peak work volumes, and low counter-related new savings accounts , Normal deposit cancellation, periodic new mouth / continuous continuation, periodic cancellation, receipt of notifications, foreign currency exchange / T / C work volume peak index, lobby corner window high, low (consultation), low (notifications) ) Visitor peak index, lobby corner automatic service machine peak index, and high, low (consultation), and low (notification) counters.

  In step S302, if the form data for (N−M + 1) months from N months before M months before the prediction target month has not been extracted from the database 102, the process returns to step S301, and the amount of work changes from the database 102. The form data necessary for predicting the number of visitors is extracted.

In step S304, with respect to the item calculated in step S303, the change rate is calculated by dividing the actual value before M months of the prediction target month by the average value from N months to M months before the calculated prediction target month. Calculate and go to step S305. For example, N is 15, M is 3, the prediction target month is December 2010, the average value of item A from September 2009 to September 2010 is 10, and the result of item A Assuming that the value is 12, the calculated rate of increase / decrease of item A is as follows.
Increase / decrease rate = 12 ÷ 10 = 1.2
In step S305, for the item calculated in step S303, the increase / decrease rate calculated in step S304 is multiplied by the actual value before K months of the prediction target month, and the multiplied value is used as the predicted value for the amount of work and the monthly amount of work. Is stored in the data 212 of the memory 202, and the process proceeds to step S306. However, K is a natural number and satisfies N>K> M. For example, N is 15, K is 12, M is 3, the prediction target month is December 2010, the increase / decrease rate of the calculated item A is 1.2, and the prediction target month Assuming that the actual value of item A in December 2009, which is 12 months ago (the same month of the previous year), is 20, the predicted value of item A to be calculated is as follows.
Predicted value = 20 × 1.2 = 24
The natural numbers N, M, and K can be set according to the actual state of business. However, when N is 15, K is 12, and M is 3, the annual increase and decrease trends and seasonal variations It is possible to calculate a predicted value in consideration of

  In step S306, it is determined whether all items necessary for prediction have been calculated. If all the items necessary for the prediction are calculated, the predicted values stored in the data 213 of the memory 202 in step S305 are output to the screen display device 205, and the process is terminated, and all the items are not calculated. In the case, the process returns to step S303.

  4 and 5 are flowcharts showing a method for predicting the transition of the amount of work and the transition of the number of visitors according to the embodiment of the present invention. The process of the prediction method when acquiring the prediction result of the business amount transition and the visitor number transition by executing the application 212 stored in the memory 202 of the prediction apparatus 101 will be described.

  When the application 212 stored in the memory 202 of the prediction device 101 is executed and a request for prediction of business volume transition and visitor number transition is received from the input device 204 of the prediction device 101, in step S401, referring to FIG. By assigning a code value to each date of the prediction target month based on a group definition described later, the dates of the prediction target month are classified into groups, and the process proceeds to step S402.

  In step S402, the form data necessary for predicting the business amount transition and the visitor number transition is extracted from the database 102, and the process proceeds to step S403. For example, the form data extracted in step S402 includes a peak day overview table.

  In step S403, each date of the form data extracted in step S402 is assigned by assigning a code value to each date of the form data extracted in step S402 based on the group definition described later with reference to FIG. The group is classified, and the process proceeds to step S404.

  In step S404, it is determined whether there is a matching code value among the code values of the form data assigned in step S403 with respect to the code value of each date of the prediction target month assigned in step S401. If there is a matching code value among the code values of the form data assigned in step S403 for the code value of each date of the prediction target month assigned in step S401, the process proceeds to step S405, and the matching code value is determined. If not, the process proceeds to step S406.

  In step S405, the peak index of each item of all dates having the same code value is extracted from the form data extracted in step S402, and the process proceeds to step S406. For example, the peak indices to be extracted are the peak indices for high counter-related deposits, withdrawals, wire transfers, document transfers, tax payments, yen exchanges, passbook entry / carry-over work volumes, and low counter-related new savings account , Normal deposit cancellation, periodic new mouth / continuous continuation, periodic cancellation, receipt of notifications, foreign currency exchange / T / C work volume peak index, lobby corner window high, low (consultation), low (notifications) ) And the peak index of the number of visitors of the lobby corner automatic service machine.

  In step S <b> 406, it is determined whether or not all the form data for (N−M + 1) months from N months before M months before the prediction target month have been extracted from the database 102. For example, if N is 15, M is 3, and the prediction target month is December 2010, all the form data for 13 months from September 2009 to September 2010 is extracted from the database 102. Determine if you did.

  In step S406, when the form data for (N-M + 1) months from N months before the prediction target month to M months has been extracted from the database 102, the process proceeds to step S501, and from N months before the prediction target month. If the form data for (N−M + 1) months up to M months ago has not been extracted from the database 102, the process returns to step S402.

  In step S501, the average value of the extracted peak index is calculated, the calculated average value is stored in the data 213 of the memory 202 as the predicted value of the peak index, and the process proceeds to step S502. For example, if N is 15, M is 3, and the prediction target month is December 2010, the code value to be described later with reference to FIG. 6 is S10, that is, December 10, 2010. In the case of (Friday), the code value S10 is assigned to the form data for 13 months from September 2009 to September 2010. Tuesday, November 10, 2009, December 10, 2009 (Thursday), February 10, 2010 (Wednesday), March 10, 2010 (Wednesday), May 10, 2010 (Monday), June 10, 2010 (Thursday), August 10, 2010 (Tuesday), which is form data for August, 2010, September 10 (Friday), and the average value of the extracted peak index is the average value of the peak index for August.

  In step S502, it is determined whether all items necessary for prediction have been calculated. When all items necessary for the prediction are calculated, the predicted value stored in the data 213 of the memory 202 in step S501 is output to the screen display device 205, the process is terminated, and all the items are not calculated. Returns to step S501.

  FIG. 6 is a diagram exemplifying groups and code values used in the method for predicting the transition of the business amount and the transition of the number of visitors according to the embodiment of the present invention. The groups are divided according to characteristics such as whether the day is a busy day of the end of the month, the 10th or 25th, the off-day, or another level day, and a code value is assigned to each of the groups.

  In steps S401 and S403 described with reference to FIGS. 4 and 5, the priority when assigning code values to dates is in the order of # 1, # 2,..., # 13, and the highest priority order. Is # 1, and the lowest priority is # 13. According to this priority order, a code value is uniquely determined for each date.

  FIG. 7 is a diagram illustrating an example in which a group and a code value used in a method for predicting a change in the amount of work and a change in the number of visitors according to an embodiment of the present invention are applied. FIG. 7 shows an example in which a code value is assigned to each date in December 2010. On December 10, 2010 (Friday), S10 is assigned, and on December 25, 2010 (Saturday). Z01 is assigned, and S31 is assigned on December 28, 2010 (Tuesday).

  FIG. 8 is a diagram exemplifying prediction results of the business amount transition and the visitor number transition according to the embodiment of the present invention. FIG. 8 shows an example where N is 15, M is 3, and the prediction target month is December 2010. Based on the form data for 13 months from September 2009 to September 2010. In this example, the predicted values calculated in the above are output as the work volume personnel, the monthly work volume, the peak index of the customer service section, and the peak index of the number of visitors in the lobby corner.

  According to the present embodiment, it is possible to calculate a predicted value of the amount of work and the number of visitors in the next month, and to provide a prediction result of the amount of work and the number of visitors.

  The present invention can be used for business amount transition prediction and visitor number transition prediction in various fields.

DESCRIPTION OF SYMBOLS 101 Prediction device 102 Database 104, 106 Computer 105 Internet 201 Central processing unit 202 Memory 203 Storage device 204 Input device 205 Screen display device 206 Network connection device 211 Basic software 212 Application 213 Data

Claims (4)

In a computer connected to a database that stores data on past work volume personnel, monthly work volume, and peak index on a monthly basis,
Extracted from the database data related to the work volume from N months to M months before the forecast month, monthly work volume, and peak index, where N and M are natural numbers and satisfy N> (M + 1) And steps to
Calculating an average value of each item of data relating to the extracted office work volume and monthly work volume;
Calculating an increase / decrease rate of each item by dividing a value of M months before the extracted prediction target month by the calculated average value;
For each item of data relating to the extracted amount of office work and monthly amount of work, the calculated increase / decrease rate is a value K months before the extracted prediction target month, where K is a natural number and N> Multiplying K> M, and storing the multiplied value as a predicted value for the work amount personnel and the monthly work amount;
Assigning code values according to the characteristics of the day to each date from N months before M months before the prediction target month and to each date of the prediction target month;
Determining whether there is a code value corresponding to a code value from N months before to M months before the assigned prediction target month for the code value of each date of the assigned prediction target month; ,
Extracting the peak index of each item on all dates for which the determined code values match from data relating to the peak index from N months to M months before the extracted prediction target month;
Calculating an average value of the peak index of each extracted item, and storing the calculated average value as a predicted value related to the peak index;
A prediction method, comprising: a step of outputting a predicted value related to the stored work volume personnel and monthly work volume, and a predicted value related to the stored peak index.   The prediction method according to claim 1, wherein the stored predicted value for the peak index includes a predicted value for the peak index for the number of visitors. It is a forecasting device for the transition of the amount of work and the number of visitors connected to the database that stores the data on the past work volume, the monthly work volume, and the peak index,
Extracted from the database data related to the work volume from N months to M months before the forecast month, monthly work volume, and peak index, where N and M are natural numbers and satisfy N> (M + 1) Means to
Means for calculating an average value of each item of data relating to the extracted office work volume and monthly work volume;
Means for calculating an increase / decrease rate of each item by dividing a value of M months before the extracted prediction target month by the calculated average value;
For each item of data relating to the extracted amount of office work and monthly amount of work, the calculated rate of change is the value K months before the extracted forecast target month, where K is a natural number and N> Means for satisfying K> M, and storing the multiplied value as a predicted value relating to the amount of work and the amount of monthly work;
Means for assigning code values according to the characteristics of the day to each date from N months to M months before the prediction target month and to each date of the prediction target month;
Means for determining whether there is a code value corresponding to a code value from N months to M months before the assigned prediction target month, for each assigned code value of the prediction target month; ,
Means for extracting the peak index of each item on all dates for which the determined code value matches, from data relating to the peak index from N months to M months before the extracted prediction target month;
Means for calculating an average value of the peak index of each extracted item, and storing the calculated average value as a predicted value for the peak index;
A prediction apparatus, comprising: means for outputting a predicted value related to the stored work volume personnel and monthly work volume, and a predicted value related to the stored peak index.   The prediction device according to claim 3, wherein the stored predicted value for the peak index includes a predicted value for the peak index for the number of visitors.

Images