Method and system for efficient segmentation for forecasting
AssigneeKinaxis Inc.
InventorBehrouz Haji Soleimani
Life of this patent
Estimating the term…
Abstract
Systems and methods for partitioning forecast items into segments, or groups, based on an attribute. Once partitioned, segments are aggregated based on a predetermined memory size limit using an aggregation method. Aggregation methods include methods for providing segments that have a similar number of records. Segmentation is further enhanced by including variance reduction gain as a metric for selecting the order of attributes.
Claims15
A computing apparatus comprising:
- a processor;
- and
- a memory storing instructions that, when executed by the processor, configure the apparatus to:
- segment a root node that includes a dataset, into a plurality of segments according to a first attribute, the root node exceeding a memory threshold;
- aggregate two or more of the segments into one or more aggregated segments, the one or more aggregated segments each having a memory size equal to or below the memory threshold;
- and
- recursively:
- segment, according to successive attributes, any segment that exceeds the memory threshold;
- and
- aggregate two or more segments into aggregated segments that are less than or equal to the memory threshold;
- until no further segmentation or aggregation can be performed.
A non-transitory computer-readable storage medium, the computer-readable storage medium including instructions that when executed by a computer, cause the computer to:
- segment a root node that includes a dataset, into a plurality of segments according to a first attribute, the root node exceeding a memory threshold;
- aggregate two or more of the segments into one or more aggregated segments, the one or more aggregated segments each having a memory size equal to or below the memory threshold;
- and
- recursively:
- segment, according to successive attributes, any segment that exceed the memory threshold;
- and
- aggregate two or more segments into aggregated segments that are less than or equal to the memory threshold,
- until no further segmentation or aggregation can be performed.
A computer-implemented method for segmenting a dataset into a plurality of segments, the dataset grouped according to a plurality of attributes, the method comprising:
- segmenting a root node that includes the dataset, into a plurality of segments according to a first attribute, the root node exceeding a memory threshold;
- aggregating two or more segments into one or more aggregated segments, the one or more aggregated segments each having a memory size equal to or below the memory threshold;
- and
- recursively:
- segmenting, according to successive attributes, any segment that exceed the memory threshold;
- and
- aggregating two or more segments into aggregated segments that are less than or equal to the memory threshold;
- until no further segmentation or aggregation can be performed.
Dependencies are inferred from the claim text.