US 2010/0138274 A1Published application

Method for determining daily weighting factors for use in forecasting daily product sales

AssigneeTeradata US, Inc.

InventorsArash Bateni, Edward Kim

Life of this patent

Estimating the term…

Loading patent…

Abstract

A product demand forecasting methodology is presented that applies daily weight values to a weekly forecast to determine daily forecasts for a product or service. The method determines daily weight values for use in forecasting current product sales by blending daily weight values calculated from historical demand data for both recent weeks and year-prior weeks. Recent weeks are used to account for recent correlations and alternation effects, and year-prior weeks are used to account for seasonality effects. The method automatically calculates a measure of significance for the daily weights calculated from the recent weeks and year-prior weeks. The significance of each week is applied as a weighting factor during the blending of recent weeks and year-prior daily weight values.

Claims8

Dependent claims

A method to determine daily weight values for use in forecasting daily demand for a product during a forecast week, said method comprising the steps of:

storing within a data warehouse historical demand data for said product;
retrieving from said data warehouse historical daily demand data for said product for a period of at least one week immediately preceding said forecast week;
for each day of said at least one immediately preceding week, calculating a daily weight value for said product by dividing the historical daily demand for a corresponding day of said at least one immediately preceding week by a total of said historical daily demands for all seven days of said at least one immediately preceding week;
retrieving from said data warehouse historical daily demand data for said product for a period of at least one week one year prior to said forecast week;
for each day of said at least one year prior week, calculating a daily weight value for said product by dividing the historical daily demand for a corresponding day of said at least one year prior week by a total of said historical daily demands for all seven days of said at least one year prior week;
and
blending said corresponding daily weight values for said at least one immediately preceding week and said at least one year prior week to determine a daily weight value for use in forecasting daily demand for a product.

4 more elements

A method to determine daily weight values for use in forecasting daily demand for a product during a forecast week, said method comprising the steps of:

storing within a data warehouse historical demand data for said product, said demand data comprising weekly and daily demand data for at least fifty-three weeks prior to said forecast week;
retrieving from said data warehouse historical daily demand data for said product for weeks one through four immediately preceding said forecast week;
for each day of said at weeks one through four, calculating a daily weight value for said product by dividing the historical daily demand for a corresponding day of said weeks one through four by a total of said historical daily demands for all seven days of said weeks one through four;
retrieving from said data warehouse historical daily demand data for said product for weeks fifty through fifty-three preceding said forecast week;
for each day of said weeks fifty through fifty-three, calculating a daily weight value for said product by dividing the historical daily demand for a corresponding day of said weeks fifty through fifty-three by a total of said historical daily demands for all seven days of said weeks fifty through fifty-three;
and
blending said corresponding daily weight values for said weeks one through four and said weeks fifty through fifty-three to determine a daily weight value for use in forecasting daily demand for a product.

4 more elements

Dependencies are inferred from the claim text.

Facts

Application: 12/326,145Filed 2 Dec 2008
Published: 3 Jun 2010
Term ends: Not estimatedThe source records don’t support an estimate.

Classification4 codes

G06Q 30/02
Commerce: marketing; Price estimation or determination; Fundraising
G06Q 10/04
Administration; Management: forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or…
G06Q 30/0202
Market modelling; Market analysis; Collecting market data: market predictions or forecasting for commercial activities
G06Q
Information and communication technology [ict] specially adapted for administrative, commercial, financial, managerial