Use of object group models and hierarchies for output predictions

Patent metadata

Field	Value
Publication number	US-2016239749-A1
Application number	US-201614987982-A
Country	US
Kind code	A1
Filing date	Jan 5, 2016
Priority date	Oct 28, 2008
Publication date	Aug 18, 2016
Grant date	—

Abstract

Computer-implemented systems and methods are provided for predicting outputs. Global output fractions associated with an object are approximated. Outputs for a group are predicted based upon a cyclical aspect component and a movement prediction. An output prediction is calculated based upon the predicted outputs for a related object group and the approximated global output fraction for a particular object.

First claim

What is claimed is:

1 . A system, comprising:
a network node in data communication with one or more remote nodes, the network node including one or more processors; and one or more non-transitory computer-readable storage mediums containing instructions configured to cause the one or more processors to perform steps including:
receiving, by the network node, past data stored in a multidimensional online analytical processing database, wherein the past data is organized according to a spatial hierarchy that includes a plurality of levels and an object hierarchy that includes a plurality of levels, wherein each level in each hierarchy includes a corresponding amount of detail, wherein the plurality of levels include one or more related object groups;
evaluating a selection of a level in the spatial hierarchy and a level in the object hierarchy, wherein the selected levels in each hierarchy have a corresponding amount of detail;
generating a cyclical aspect component using past data located at the selected levels in each hierarchy;
evaluating a selection of a different level in the spatial hierarchy and a different level in the object hierarchy, wherein the different levels in each hierarchy have a greater corresponding amount of detail;
generating a movement component using past data located at the different levels in each hierarchy;
generating a base requirement component for a related object group in the plurality of levels using the cyclical aspect component and the movement component;
generating an individual approximated global output fraction for a member of a related object group using the past object data and a global output fraction model, wherein the individual approximated global output fraction is a proportion of total outputs for the related object group expected for a particular object; and
predicting approximated output for the particular object using the base requirement component and the individual approximated global output fraction for the particular object, wherein predicting includes multiplying the base requirement component by the individual approximated global output fraction.

Assignees

• Sas Inst Inc

Inventors

• Peredriy Sergiy
• Chien Yung-Hsin
• Chaudhuri Arin
• Mcguirk Ann Mary
• Xiao Yongqiao

Classifications

• G06Q10/04 Primary

  Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202) · CPC title

• G06F17/30592

  Physics · mapped topic

• G06N5/048 Primary

  Fuzzy inferencing · CPC title

• G06Q30/0202

  Market predictions or forecasting for commercial activities · CPC title

• G06F16/283

  Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP · CPC title