Title:
COMPOSITE MACHINE LEARNING ARCHITECTURE FOR INFERENCE BASED ON MULTIPLE SUBSIDIARY MODELS
Document Type and Number:
WIPO Patent Application WO/2024/059288
Kind Code:
A3
Abstract:
Architectures for machine inference are provided. A plurality of subsidiary models are instantiated. For each of the plurality of subsidiary models, at least one feature vector selected from a plurality of data is provided as input. A respective output is received from each of the plurality of subsidiary models. The respective output of each of the plurality of subsidiary models is provided to a super model. A score is received from the super model.
Inventors:
OLIVEIRA ALINE (US)
Application Number:
PCT/US2023/032907
Publication Date:
April 25, 2024
Filing Date:
September 15, 2023
Export Citation:
Assignee:
TRAIVE INC (US)
International Classes:
G06N3/042; G06N3/02; G06N3/045; G06N20/00; G06Q30/018
Attorney, Agent or Firm:
HUESTIS, Erik, A. et al. (US)
Download PDF: