Back to results
US 11,537,598 B1Granted
Effective ensemble model prediction system
AssigneeInternational Business Machines Corporation
InventorsDong Hai Yu, Song Bo, Jun Wang, Jiang Bo Kang, Yao Dong Liu
Life of this patent
Estimating the term…
Abstract
A method for instantiating a first specific model, as one record, based on a model scoring template, clustering a plurality of models into a plurality of groups to obtain a plurality of clusters, computing a model contribution weight value for the given cluster, choosing a set of top model(s) from the plurality of models, and updating the model(s) of the set of top model(s) dynamically.
Claims15
A computer-implemented method (CIM) comprising:
- instantiating a first specific model, as one record, based on a model scoring template;
- clustering a plurality of models into a plurality of groups to obtain a plurality of clusters;
- for each given cluster of the plurality of clusters, computing a model contribution weight value for the given cluster;
- choosing a set of top model(s) from the plurality of models;
- updating the model(s) of the set of top model(s) dynamically;
- using predictive model markup language (PMML) to recover a plurality of targeted predicted values using predictive model markup language (PMML);
- and
- recovering the first specific model from a plurality of targeted predicted values.
5 more elements
A computer program product (CPP) comprising:
- a set of storage device(s);
- and
- computer code stored collectively in the set of storage device(s), with the computer code including data and instructions to cause a processor(s) set to perform at least the following operations:
- instantiating a first specific model, as one record, based on a model scoring template,
- clustering a plurality of models into a plurality of groups to obtain a plurality of clusters,
- for each given cluster of the plurality of clusters, computing a model contribution weight value for the given cluster,
- choosing a set of top model(s) from the plurality of models,
- updating the model(s) of the set of top model(s) dynamically,
- using predictive model markup language (PMML) to recover a plurality of targeted predicted values using predictive model markup language (PMML), and
- recovering the first specific model from a plurality of targeted predicted values.
7 more elements
A computer system (CS) comprising:
- a processor(s) set;
- a set of storage device(s);
- and
- computer code stored collectively in the set of storage device(s), with the computer code including data and instructions to cause the processor(s) set to perform at least the following operations:
- instantiating a first specific model, as one record, based on a model scoring template,
- clustering a plurality of models into a plurality of groups to obtain a plurality of clusters,
- for each given cluster of the plurality of clusters, computing a model contribution weight value for the given cluster,
- choosing a set of top model(s) from the plurality of models, and
- updating the model(s) of the set of top model(s) dynamically,
- using predictive model markup language (PMML) to recover a plurality of targeted predicted values using predictive model markup language (PMML), and
- recovering the first specific model from a plurality of targeted predicted values.
8 more elements
Dependencies are inferred from the claim text.