METHOD FOR TESTING A MACHINE LEARNING MODEL BASED ON AN ANALYSIS OF A CLASS PREDICTION VALUE

FR3149112A1Pending Publication Date: 2024-11-29GISKARD AI
2 Cites 0 Cited by

Patent Information

Application Number
FR2023005156
Authority / Receiving Office
FR · FR
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-24
Publication Date
2024-11-29
Patent Text Reader

Abstract

METHOD FOR TESTING A MACHINE LEARNING MODEL BASED ON AN ANALYSIS OF A CLASS PREDICTION VALUE. A method for testing a machine learning model comprising: acquisition of a first machine learning model (ML1); acquisition (ACQ1) of a dataset (DS1) describing a set of digital objects ({OBp}), each digital object (OBi) comprising a set of descriptors ({Fik}); said graphical representation displaying an activatable symbol; activation (ACT1) of said symbol and generation of a test of the first model (ML1), said test comprising: extraction of a subset of digital objects ({OBp'}) whose first-class prediction value (CL1) meets the same condition as the first-class prediction value (CL1) of the first object (OBi); generation of an anomaly indicator (IND2). FIG. 1
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Method and system for testing machine learning models

    US20220067589A1

  • Software model testing for machine learning models

    US20230036289A1