JP6844304B2 - Demand forecast model evaluation device, demand forecast model evaluation method and demand forecast model evaluation program - Google Patents

Description

The present invention relates to a demand forecast model evaluation device for evaluating a demand forecast model suitable for evaluating the usefulness of a built demand forecast model in actual business , a demand forecast model evaluation method, and a demand forecast model evaluation program .

Patent Document 1 discloses a demand forecast model evaluation method. In this demand forecast model evaluation method, with respect to the actual demand value and the forecasted demand value in the evaluation period of the demand forecast model, the deviation value between the actual demand value and the forecasted demand value is calculated for each supply cycle of the product, and the deviation value is calculated. Oversupply costs and undersupply costs are calculated according to the values and presented as a result. In this demand forecast model evaluation method, as the usefulness of the constructed forecast model for actual business, the forecast model is evaluated by regarding the deviation value between the actual demand value and the demand forecast value as the theoretical inventory.

Japanese Unexamined Patent Publication No. 10-228463

However, in the demand forecast model evaluation method described in Patent Document 1, the deviation value between the actual demand value and the predicted demand value is regarded as the theoretical inventory. Therefore, in the conventional method, the oversupply cost and the undersupply cost are hypothetical evaluation values and do not show the transition of the actual inventory quantity. Therefore, it is unclear how the actual inventory quantity changes even if it is presented to the user as a time-series transition diagram of the theoretical inventory from the start time to the end time of the prediction model evaluation period by the conventional method.

In addition, although it is possible to calculate the difference between the predicted value and the actual demand (error value) for each prediction target using MAD, RMSE, MAPE, etc., which are generally used to calculate the accuracy of prediction, the error value is calculated for each of a large number of prediction targets. It is not realistic to confirm, and even if the error value is small, it cannot be said that the prediction is easy for the user to understand.

Furthermore, even if the inventory quantity gradually decreases in the actual period after the introduction of the conventional demand forecast model system, the gradual decrease change "appropriately suppresses orders based on the demand forecast value and the inventory quantity. Wasn't it possible, but did the inventory decrease due to the increase in demand? "Or" Did the inventory decrease due to the appropriate restraint of orders based on the forecast value and the number of inventories? " Is unknown, and the inventory reduction effect of the demand forecast model is unknown.

In addition, it is impossible to calculate the inventory reduction effect before the introduction of the system for the evaluation of the system in the actual working state of the built demand forecast model. Therefore, there is a problem that it is not possible to estimate how long the forecast system introduction cost can be recovered before the system is introduced.

The present invention has been made in view of the above problems of the prior art, and is a demand forecast model evaluation device , a demand forecast model evaluation method, and a demand that can confirm how the forecast stock quantity of goods changes by the demand forecast model. The purpose is to provide a predictive model evaluation program.

The demand forecast model evaluation device of the present invention is a demand forecast model evaluation device that evaluates a demand forecast model, and is the actual number of goods delivered in the reference period of the past inventory quantity management period and the evaluation period after the reference period. Based on the first reception unit that accepts the input up to the final inventory actual value in the above, the second reception unit that accepts the input of the expected delivery value of the goods by the demand forecast model in the reference period, and the number of inventories and the forecasted issue value. A forecast order quantity calculation unit that calculates the forecast order quantity for each unit period cycle, and an estimated inventory quantity calculation unit that calculates the estimated inventory quantity for each unit period cycle based on the actual number of goods issued and the forecast order quantity. An inventory evaluation unit that calculates the difference between the estimated inventory quantity at the end of the evaluation period and the final inventory actual value based on the estimated inventory quantity at the end of the evaluation period and the final inventory actual value, and the above. It is characterized by having a display unit for displaying the difference.

The demand forecast model evaluation method of the present invention is a demand forecast model evaluation method for evaluating a demand forecast model, in which the computer device determines the actual number of goods delivered in the reference period of the past inventory quantity management period and the reference period. The first reception step that accepts the input up to the final inventory actual value in the subsequent evaluation period, the second reception step that accepts the input of the delivery forecast value of the article by the demand forecast model in the reference period, the inventory quantity and the delivery. The forecast order quantity calculation step that calculates the forecast order quantity for each unit period cycle based on the forecast value, and the estimated inventory quantity that calculates the estimated inventory quantity for each unit period cycle based on the actual number of goods issued and the forecast order quantity. Inventory valuation that calculates the difference between the estimated inventory quantity at the end of the evaluation period and the final inventory actual value based on the calculation step and the estimated inventory quantity at the end of the evaluation period and the final inventory actual value. It is characterized in that a step and a display step for displaying the difference are executed.
The demand forecast model evaluation program of the present invention is a demand forecast model evaluation program that evaluates a demand forecast model, and the actual number of goods delivered to the computer device in the reference period of the past inventory quantity management period and the reference period. The first reception step that accepts the input up to the final inventory actual value in the subsequent evaluation period, the second reception step that accepts the input of the delivery forecast value of the article by the demand forecast model in the reference period, the inventory quantity and the delivery. The forecast order quantity calculation step that calculates the forecast order quantity for each unit period cycle based on the forecast value, and the estimated inventory quantity that calculates the estimated inventory quantity for each unit period cycle based on the actual number of goods issued and the forecast order quantity. Inventory valuation that calculates the difference between the estimated inventory quantity at the end of the evaluation period and the final inventory actual value based on the calculation step and the estimated inventory quantity at the end of the evaluation period and the final inventory actual value. It is characterized in that a step and a display step for displaying the difference are executed.

It is a block diagram which shows the whole structure of the demand forecast model evaluation system including the demand forecast model evaluation apparatus of the Example by this invention. It is a plurality of tables showing an example of the image of the data stored in the storage unit as the acquired storage file in the demand forecast model evaluation device of the embodiment according to the present invention. It is a flowchart which shows the operation of the demand forecast model evaluation apparatus of the Example by this invention. A table (Fig. 4A) showing changes in both the predicted inventory quantity and the actual inventory quantity obtained by the demand forecast model evaluation device of the embodiment according to the present invention, and the evaluation period, the number of target products, the introduction effect amount, and the inventory quantity. It is an example (FIG. 4B) of the display result of increase / decrease. It is a graph which showed the transition of both the predicted stock quantity and the actual stock quantity of FIG. 4A.

Hereinafter, the demand forecast model evaluation device and the demand forecast model evaluation method according to the embodiment of the present invention will be described in detail with reference to the drawings. The present invention is not limited to the following examples.

[Demand forecast model evaluation device]
FIG. 1 shows the overall configuration of a demand forecast model evaluation system 103 including a demand forecast model evaluation device 100 of an embodiment, a wireless communication terminal 101 connected to the device, and a wired communication terminal 101A. The demand forecast model evaluation device 100 can be realized by, for example, a computer device 109.

An example of the hardware configuration of the demand forecast model evaluation device 100 in FIG. 1 is a computer device 109. Further, the demand forecast model evaluation device 100 includes a hardware network connection unit (NIC: network interface card) 119. The network connection unit transmits various data to the communication terminals 101 and 101A, and receives various data from the communication terminal 101. The communication terminals 101 and 101A are a personal computer (personal computer), a tablet terminal, or the like having a communication function used by the user to evaluate the demand forecast model, and are a mobile phone network or an Internet network with respect to the demand forecast model evaluation device 100. Etc. are connected via a communication network IP including the above. In this embodiment, the demand forecast model evaluation device 100 can be used by a plurality of users by being connected to a large number of two or more communication terminals. Further, these communication terminals 101 and 101A have a function of acquiring an e-mail destination address at the input unit and transmitting an e-mail to the output unit, similarly to the demand forecast model evaluation device 100.

In the computer device 109 shown in FIG. 1, the CPU 111 uses the RAM 115 as a primary storage work memory according to a program stored in a storage unit 116 composed of a ROM 114 and / or a hard disk device, an SSD (solid state drive), or the like. And control the whole system. Further, the CPU 111 executes the demand forecast model evaluation method according to the embodiment based on the program stored in the storage unit 116 according to the user's instruction input via the mouse 112a or the keyboard 112. The above components are connected to each other via a bus bus. A display device 113 such as an LCD connected to the computer device 109 displays an input standby screen for a process to be described later executed by the CPU 111, a process progress, a process result, a search result, and the like.

The storage unit 116 of the computer device 109 stores a demand forecast model evaluation program 117 that functions as the demand forecast model evaluation device 100. The demand forecast model program to be evaluated may be installed in the storage unit 116. The CPU 111 reads the demand forecast model evaluation program 117 from the storage unit 116, deploys it to the RAM 115 (storage unit 116 if necessary), and sequentially executes the process included in the demand forecast model evaluation program 117. The demand forecast model evaluation program 117 is a program that causes the computer device 109 to function as the first reception unit 120, the second reception unit 121, the forecast order quantity calculation unit 122, the estimated inventory quantity calculation unit 123, and the inventory evaluation unit 124. is there.

The first reception unit 120 as the input unit and the predicted order quantity calculation unit 122 acquire the data of the evaluation target period (for example, the past inventory quantity management period) and store these as a storage file in the storage unit 116. Here, the data is the product ID, year / month, number of shipments :, number of unstocked items :, unit price (yen), number of safety items: A, lead time: k (adjustment of number of orders). It refers to the period from ordering to acquisition (generally, the sum of the ordering cycle and procurement lead time) including the period, the predicted value Y, and the like.

Before the introduction of the demand forecast model system, the data other than the forecast value is generally the information stored in the core database DB, and therefore, for example, it is acquired by connecting to the core database DB via the communication network IP. If it is difficult to connect to the core database DB, the relevant data may be acquired in CSV format via another medium. The forecast value is calculated by the demand forecast model program to be evaluated by using the time series forecast or the like.

FIG. 2 shows an example of a tabular image of data acquired and stored as a storage file in the storage unit. If the necessary items can be collected from the core database DB or the like, the necessary items may not exist on the same table as shown in FIG. In addition, since the data item name differs for each company, it is sufficient if there is applicable data.

(Explanation of data)
As shown in FIG. 2, the lead time based on the period from the ordering of the product to the arrival of the product is stored. The lead time is set as the period from the ordering of the product to the arrival of the product, such as the transportation time of the product, the production constraint time (including the preparation, manufacturing, and packing time), and the paperwork time. The lead time may be stored for each product by inputting an appropriate value into the device. In this embodiment, the unit period (i) is set to the monthly unit, but the unit period (i) is not limited to this, and may be set to the daily unit or the hour unit such as one day or one hour.

The number of inventories is calculated based on the number of inventories in the previous month, the number of shipments in the current month, and the number of unstocked items, and is stored in each unit period (every month in this embodiment). The stock quantity Z is defined by the following formula.
Z [i] = Z [i-1] + M [i] -J [i]
During the ceremony
i: The number of unit periods.
J [i]: The actual number of shipments (demand) in the unit period i.
H [i]: The actual order quantity in the unit period i.
M [i] = H [ik]: The number of undelivered products in the unit period i (the number of undelivered products in the ordered quantity before the unit period i to the k period (lead time)). For example, if it is currently May [i], 10 pieces will be ordered (H [i-2]) in March [i-2] (H [i-2] = 10) and will be delivered in May. If there is M [i] that has not arrived or is not ready for use, M [i] = 10. Therefore, assuming that the last day of May [i] is present, M [i] is April [i] because there is an unstocked number (M [i]) on the last day of April [i-1]. -1] is added to the stock quantity (Z [i-1]), and the shipment quantity (J [i]) in May [i] is subtracted to obtain the stock quantity Z [i] in May [i]. Become.

The forecast required quantity R, which is data for avoiding inventory shortage, is defined by the following formula.
R [i] = Y [i] + A [i]
During the ceremony
i: The number of unit periods.
Y [i]: A shipping forecast value for the unit period i (this is obtained from the forecast model to be evaluated).
A [i]: The number of safe stocks in the unit period i (this is a stock for preventing shortages to some extent due to uncertain factors such as demand fluctuations, and is preferably 0 or more).

The reference period is a period that serves as a reference for simulation based on the estimated shipment value from the forecast model to be evaluated. In this embodiment, max (k), which has the longest k period (lead time) among the products to be predicted, is set as the reference period. The reference period is not limited to this, and may be set to a period shorter than the evaluation target period (past inventory quantity management period).

In this embodiment, the first reception unit 120 in FIG. 1 accepts inputs up to the actual number of goods shipped in the reference period of the evaluation target period and the final inventory actual value in the evaluation period after the reference period, and the second reception unit 121 is configured to accept the input of the forecast value of the shipment of goods by the demand forecast model in the reference period.

[Demand forecast model evaluation method]
FIG. 3 is a flowchart showing the operation of the demand forecast model evaluation device by the demand forecast model evaluation method of this embodiment. Using the data acquired by the first reception unit 120 and the second reception unit 121, the inventory quantity when an order is placed using the shipping forecast value from the forecast model to be evaluated is calculated and compared with the actual inventory quantity.

First, the first reception unit 120 sets the product ID, year / month, number of shipments, number of stocks, number of unstocked items, unit price, number of safe stocks and lead time data, the actual number of goods shipped in the above reference period, and the final in the evaluation period. It accepts inputs up to the actual inventory value and stores them in the storage unit 116 (step S1). Next, the second reception unit 121 receives the above-mentioned shipment forecast value and forecast required quantity data and stores them in the storage unit (step S2).

Next, as a loop for the product ID, the initial value I of the product ID is set to 1, and the evaluation period loop processing of the following is repeated (increment 1) up to the number of target products I = 9000 (step S3). Then, as the evaluation period loop, the initial value i = 1 of the number of evaluation periods is set, and the following processing is repeated (increment 1) until the number of evaluation periods i = N + max (k) (step S4). Here, N is the number of evaluation periods for evaluating the effect of the predicted model, and max (k) is the period in which the k period (lead time) is the longest. In this embodiment, max (k) is used as the reference period.

Next, the predicted order quantity Ho is calculated based on the forecast required quantity, the inventory at the forecast time point, and the unstocked remaining quantity, and the result is stored in the storage unit (step S5). The estimated order quantity Ho is defined by the following formula.
Ho = R [i]-(Z'[i-1] + M [i])
During the ceremony
i: The number of unit periods.
R [i]: The predicted required quantity of the unit period i (= Y [i] + A [i]).
Z'[i-1]: The inventory quantity at the end of the previous month in the forecast unit period i.
M [i] = H [ik]: The actual number of undelivered items in the unit period i (the number of undelivered products in the order quantity before the unit period i to the k period (lead time)).

The first forecast order quantity Ho is calculated based on the actual inventory value (inventory quantity in the unit period within the reference period) and the shipment forecast value in the cycle before the start of evaluation, and the forecast order quantity from the next time onward is one cycle. It is calculated based on the previous estimated inventory quantity (described later) and the estimated shipping value.

Next, it is determined whether or not the predicted order quantity Ho is zero or more (step S6).

Next, if the predicted order quantity Ho is zero or more (step S7), the predicted order quantity H'is set to Ho. That is,
If the required quantity R using the predicted value> (stock quantity Z'at the end of the previous month + unstocked quantity M),
Let these differences be the predicted number of orders H'.

If the predicted number of orders Ho is less than zero (step S7), the predicted number of orders H'is set to zero (step S8). The above steps S5 to S8 correspond to the forecast order quantity calculation unit 122, which calculates the forecast order quantity for each unit period cycle based on the actual unstocked remaining quantity and the shipment forecast value.

Next, when the predicted order quantity H'is used, the unstocked remaining quantity is set as M'[i] as H'[ik] (step S9).

Next, the estimated inventory quantity Z'[i] when the predicted order quantity H'is used is calculated (step S10). The estimated inventory quantity Z'[i] is defined by the following formula.
Z'[i] = Z'[i-1] + M'[i] -J [i]
During the ceremony
i: The number of unit periods.
Z'[i-1]: The inventory quantity at the end of the previous month in the forecast unit period i.
M'[i] = H'[ik]: The predicted undelivered number of units in the unit period i (the number of undelivered products in the order quantity before the unit period i to the k period (lead time)).
J [i]: The actual number of shipments (demand) in the unit period i.

The above steps S9 to S10 correspond to the estimated inventory quantity calculation unit 123, which calculates the estimated inventory quantity for each unit period cycle based on the actual shipment quantity and the forecast order quantity.

Next, the above evaluation period loop is repeated until the initial value i = 1 of the number of evaluation periods and the number of evaluation periods i = N + max (k) (step S11), and then the evaluation period loop ends.

Then, the difference between the final estimated inventory quantity and the final inventory actual value is calculated based on the estimated inventory quantity and the final inventory actual value at the end of the evaluation period (step S12). The difference D is calculated by the following formula.
D = Z'[N + max (k)]-Z [N + max (k)]
During the ceremony
Z'[N + max (k)]: The estimated number of stocks in the evaluation period N + max (k).
Z [N + max (k)]: The actual number of stocks in the evaluation period N + max (k).

The above step S12 corresponds to the inventory evaluation unit 124, which calculates the difference between the final estimated inventory quantity and the final inventory actual value based on the estimated inventory quantity and the final inventory actual value at the end of the evaluation period. ..

Next, the prediction effect (amount) is calculated based on the difference D between the final estimated inventory quantity and the final inventory actual value (step S13). That is, the prediction effect (amount) of the product (Ttl) between the difference D and the unit price of the target product is calculated.

Next, the above loop for the product ID is set to the initial value I = 1 of the product ID, and is repeated until the number of target products I = 9000 (step S14), and then the evaluation period loop ends.

Next, the predicted effect (total quantity, total amount) over the target product is calculated, and the result is stored in the storage unit (step S15). In this embodiment, the expected inventory reduction effect is calculated on the assumption that the actual inventory quantity for each product is a positive value, but the actual inventory quantity for each product is a negative value, that is, the inventory shortage quantity is In some cases, it can be calculated as the expected inventory shortage solving effect. If the actual inventory quantity of each product is a mixture of positive and negative values, the actual inventory quantity of each product is divided into positive value products and negative value products. By calculating the predicted effect, it is possible to confirm the predicted inventory reduction effect and the predicted inventory shortage solving effect, respectively.

According to the above embodiment, the display device 113, which is a display unit, displays the estimated inventory quantity for each unit period cycle of step S10, and simultaneously displays the corresponding actual shipment quantity in chronological order. The total amount of the product of the difference in S10 and the unit price of the article can be displayed as the amount calculated over all types of the article. As described above, according to the present embodiment, it is possible to display the actual inventory quantity and inventory value as a result of comparing the inventory transition and the actual inventory when the order is placed using the shipping forecast value. For example, in FIG. 4, the forecast inventory quantity and the actual inventory quantity are both 300, the reference period is set to max (k) = 3, and the evaluation period is set to N = 12 by the demand forecast model evaluation device of the embodiment according to the present invention. It is a table (FIG. 4A) showing the transition of both the predicted inventory quantity and the actual inventory quantity, and an example of the display result of the evaluation period, the number of target products, the introduction effect amount, and the increase / decrease in the inventory quantity (FIG. 4B). Further, FIG. 5 is a graph showing changes in both the predicted inventory quantity and the actual inventory quantity in FIG. 4A.

According to the present invention, the difference between the actual shipment value and the estimated shipment value is not evaluated assuming that the theoretical inventory is used, but the estimated shipment value and the actual inventory quantity, the calculated inventory quantity, and (if necessary, the remaining inventory). By determining the monthly order quantity using the number), it is possible to accurately confirm how the monthly inventory quantity changes during the evaluation period (for example, 12 months) by simulation. Furthermore, by summing the product of the inventory quantity for the evaluation period and the unit price, the inventory difference amount for all the forecasted items can be calculated and the effect amount can be confirmed. In the examples, "shipment" to be put on the market for selling a product such as a retail business is described, but the present invention is not limited thereto. The articles that can be handled in the present invention include articles other than those for sale such as replacement parts or repair parts such as products to be maintained (for example, ATM), and the present invention includes "delivery" of articles including "shipment". Can also be applied by using the "actual issue value", "number of issues", "estimated issue value", and the like.

Further, according to the invention, the effect calculation can be estimated before the introduction of the demand forecasting system by using the existing information necessary for the demand forecasting, and thereby, it is possible to estimate how much the cost can be recovered before the introduction of the demand forecasting system. It will be possible. For example, if you want to recover the investment amount in 3 years, you can estimate by the forecast effect for 1 year x 3.

Further, according to the invention, since the demand forecast model evaluation device 100 has a function of sending and receiving e-mails, if there are many products to be predicted and the evaluation period is long, it may take time to calculate the result. By automatically sending the results by e-mail to the parties concerned, it is possible to save the trouble of manually sending the evaluation results.

[Application example]
In the above embodiment, the effect of introducing the prediction system was calculated by comparing Z: the actual inventory quantity and Z': the inventory of the result of ordering using the predicted value, but after the system is introduced, Z: the actual inventory quantity. Is the number of inventories when an order is placed by the conventional method, and the effect of introducing the system can be calculated periodically by comparing the conventional method with the prediction result.

[Modification example]
If the mail sending function of the output unit is changed to the function of deploying to AppServer, all the persons who know the URL can refer to the result by referring to the predetermined URL from the Web browser. If all the evaluation results calculated by changing the evaluation period are stored, it is possible to compare a plurality of evaluation results with different periods. For example, in 2013, 2014, and 2015, it is possible to compare each year.

100 Demand Forecast Model Evaluation Device 111 CPU
115 RAM
116 Storage unit 117 Demand forecast model evaluation program 120 1st reception unit 121 2nd reception unit 122 Predicted order quantity calculation unit 123 Estimated inventory quantity calculation unit 124 Inventory evaluation unit

Claims (6)

It is a demand forecast model evaluation device that evaluates a demand forecast model.
The first reception section that accepts inputs up to the actual number of goods shipped in the reference period of the past inventory quantity management period and the final inventory actual value in the evaluation period after the reference period.
A second reception unit that accepts the input of the forecast value of the goods issued by the demand forecast model in the reference period, and
A forecast order quantity calculation unit that calculates the forecast order quantity for each unit period cycle based on the inventory quantity and the issue forecast value, and
An estimated inventory quantity calculation unit that calculates an estimated inventory quantity for each unit period cycle based on the actual number of goods issued and the estimated number of orders.
An inventory evaluation unit that calculates the difference between the estimated inventory quantity at the end of the evaluation period and the final inventory actual value based on the estimated inventory quantity at the end of the evaluation period and the final inventory actual value.
A display unit that displays the difference and
A demand forecast model evaluation device characterized by having. The demand forecast model evaluation device according to claim 1, wherein the display unit displays the estimated inventory quantity for each unit period cycle. The demand forecast model evaluation device according to claim 1 or 2, wherein the display unit displays the corresponding actual number of goods issued for each unit period cycle in chronological order. The inventory evaluation unit calculates the total of the product of the difference and the unit price of the goods over all types of the goods, and the display unit displays the calculated amount. The demand forecast model evaluation device described in. It is a demand forecast model evaluation method that evaluates the demand forecast model.
The first reception step in which the computer device accepts the input of the actual number of goods delivered in the reference period of the past inventory quantity management period and the final inventory actual value in the evaluation period after the reference period.
The second reception step of accepting the input of the delivery forecast value of the goods by the demand forecast model in the reference period, and
A forecast order quantity calculation step that calculates the forecast order quantity for each unit period cycle based on the inventory quantity and the issue forecast value, and
An estimated inventory quantity calculation step for calculating an estimated inventory quantity for each unit period cycle based on the actual number of goods issued and the predicted order quantity, and
An inventory valuation step for calculating the difference between the estimated inventory quantity at the end of the valuation period and the final inventory actual value based on the estimated inventory quantity at the end of the valuation period and the final inventory actual value.
A demand forecast model evaluation method characterized by executing a display step for displaying the difference. A demand forecast model evaluation program that evaluates a demand forecast model.
The first reception step of accepting input to the computer device up to the actual number of goods shipped in the reference period of the past inventory quantity management period and the final inventory actual value in the evaluation period after the reference period, and
The second reception step of accepting the input of the delivery forecast value of the goods by the demand forecast model in the reference period, and
A forecast order quantity calculation step that calculates the forecast order quantity for each unit period cycle based on the inventory quantity and the issue forecast value, and
An estimated inventory quantity calculation step for calculating an estimated inventory quantity for each unit period cycle based on the actual number of goods issued and the predicted order quantity, and
An inventory valuation step for calculating the difference between the estimated inventory quantity at the end of the valuation period and the final inventory actual value based on the estimated inventory quantity at the end of the valuation period and the final inventory actual value.
A demand forecast model evaluation program characterized by executing a display step for displaying the difference.

Images