Confidence measure for a deployed machine learning model
AssigneeKoninklijke Philips N.V.
InventorsMATTHIAS LENGA, RAFAEL WIEMKER, TOBIAS KLINDER, MARTEN BERGTHOLDT, HEIKE CAROLUS
Life of this patent
Estimating the term…
Abstract
Presented are concepts for obtaining a confidence measure for a machine learning model. One such concept process input data with the machine learning model to generate a primary result. It also generate a plurality of modified instances of the input data and processes the plurality of modified instances of the input data with the machine learning model to generate a respective plurality of secondary results. A confidence measure relating to the primary result is determined based on the secondary results.
Claims11
A method for obtaining a confidence measure for a machine learning model, the method comprising:
- processing input data with the machine learning model to generate a primary result;
- generating a plurality of modified instances of the input data;
- processing the plurality of modified instances of the input data with the machine learning model to generate a respective plurality of secondary results;
- and
- determining a confidence measure relating to the primary result based on the secondary results,
- wherein determining a confidence measure comprises:
- determining a measure of distribution or variance of the secondary results;
- and
- determining a confidence measure based on the determined measure of distribution or variance,
- wherein determining the measure of distribution or variance of the secondary results comprises determining at least one of:
- the inverse variance of the secondary results;
- the shannon entropy of the secondary results;
- the gini coefficient of the secondary results;
- the Kullback-Liebler divergence of the secondary results;
- and
- a concentration measure of the secondary results,
- wherein generating a plurality of modified instances of the input data comprises:
- applying a first spatial warping transformation to the input data to generate a first modified instance of the input data,
- applying a first inverse spatial warping transformation to the secondary result generated for the first modified instance of the input data.
2-5. (canceled)
A non-transitory computer readable medium storing instructions for obtaining a confidence measure for a machine learning model that, when executed by one or more processors, causes the one or more processors to:
- process input data with the machine learning model to generate a primary result;
- generate a plurality of modified instances of the input data;
- process the plurality of modified instances of the input data with the machine learning model to generate a respective plurality of secondary results;
- and
- determine a confidence measure relating to the primary result based on the secondary results, wherein determining a confidence measure comprises:
- determining a measure of distribution or variance of the secondary results;
- and
- determining a confidence measure based on the determined measure of distribution or variance,
- wherein determining the measure of distribution or variance of the secondary results comprises determining at least one of:
- the inverse variance of the secondary results;
- the shannon entropy of the secondary results;
- the gini coefficient of the secondary results;
- the Kullback-Liebler divergence of the secondary results;
- and
- a concentration measure of the secondary results,
- wherein generating a plurality of modified instances of the input data comprises:
- applying a first spatial warping transformation to the input data to generate a first modified instance of the input data,
- applying a first inverse spatial warping transformation to the secondary result generated for the first modified instance of the input data.
(canceled)
A system for obtaining a confidence measure for a machine learning model, the system comprising:
- an input interface configured to obtain input data;
- a data modification component configured to generate a plurality of modified instances of the input data;
- a machine learning model interface configured to communicate the input data and the plurality of modified instances of the input data to a machine learning model and further configured to receive a primary result generated by the machine learning model processing the input data and to receive a plurality of secondary results generated by the machine learning model processing the respective plurality of modified instances of the input data;
- and
- an analysis component configured to determine a confidence measure relating to the primary result based on the secondary results, wherein the data modification component is configured to apply a first spatial warping transformation to the input data to generate a first modified instance of the input data, wherein the analysis component is configured to determine a measure of distribution or variance of the secondary results and to determine a confidence measure based on the determined measure of distribution or variance, wherein determining the measure of distribution or variance of the secondary results comprises determining at least one of:
- the inverse variance of the secondary results;
- the shannon entropy of the secondary results;
- the gini coefficient of the secondary results;
- the Kullback-Liebler divergence of the secondary results;
- and
- a concentration measure of the secondary results, wherein the data modification component is further configured to apply a first inverse spatial warping transformation to the secondary result generated for the first modified instance of the input data.
14-15. (canceled)
Dependencies are inferred from the claim text.