DEMAND PREDICTION DEVICE, DEMAND PREDICTION METHOD AND PROGRAM

PROBLEM TO BE SOLVED: To predict the demand amount of each store with high accuracy even about an item whose sales amount is very small.SOLUTION: A demand prediction device includes a sales result storage part for storing a sales result of an item of each store every date, a whole stores prediction...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: KATO MOTOHIDE, NAKAGAWA KEIICHIRO
Format: Patent
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:PROBLEM TO BE SOLVED: To predict the demand amount of each store with high accuracy even about an item whose sales amount is very small.SOLUTION: A demand prediction device includes a sales result storage part for storing a sales result of an item of each store every date, a whole stores prediction calculation part for calculating a whole stores predicted demand amount being a predicted demand amount in the whole stores on the basis of sales results of the item read from the sales result storage part, a store-by-store prediction calculation part for calculating a store-by-store predicted demand amount being a predicted demand amount in each store on the basis of the sales result of the item read from the sales result storage part, and a store-by-store rebate prediction calculation part for correcting the store-by-store predicted demand amounts such that the total value of store-by-store predicted demand amounts is equal to the whole stores predicted demand amount. The store-by-store rebate prediction calculation part corrects a store-by-store predicted demand amount of the next day on the basis of a modified predicted value obtained by adding a carry-over value to a store-by-store predicted demand amount of the next day calculated by the store-by-store prediction calculation part with a value obtained by subtracting the corrected store-by-store predicted demand amount of the day from the store-by-store predicted demand amount of the day as the carry-over value.