US 12,282,854 B2
Batch selection policies for training machine learning models using active learning
Michael Bailey, Toronto (CA); Ziv Bar-Joseph, Cambridge, MA (US); Sven Jager, Frankfurt am Main (DE); Ruijiang Li, Shanghai (CN); and Saeed Moayedpour, Cambridge, MA (US)
Assigned to Sanofi, Paris (FR)
Filed by Sanofi, Paris (FR)
Filed on May 3, 2024, as Appl. No. 18/654,681.
Application 18/654,681 is a continuation of application No. 18/640,851, filed on Apr. 19, 2024.
Claims priority of provisional application 63/461,175, filed on Apr. 21, 2023.
Claims priority of application No. 23315451 (EP), filed on Dec. 13, 2023.
Prior Publication US 2024/0354655 A1, Oct. 24, 2024
Int. Cl. G06N 3/082 (2023.01); G06N 3/045 (2023.01); G06N 3/091 (2023.01); G06N 20/00 (2019.01)
CPC G06N 3/082 (2013.01) [G06N 3/045 (2023.01); G06N 3/091 (2023.01); G06N 20/00 (2019.01)] 30 Claims
OG exemplary drawing
 
1. A method performed by one or more computers, the method comprising:
training a machine learning model over a sequence of training iterations, comprising, at each of a plurality of training iterations in the sequence of training iterations:
selecting a current batch of model inputs for training the machine learning model at the training iteration, wherein the current batch of model inputs comprises a plurality of model inputs, wherein selecting the current batch of model inputs comprises:
generating a set of candidate batches of model inputs;
generating, for each candidate batch of model inputs, a respective score for the candidate batch of model inputs that characterizes:
(i) an uncertainty of the machine learning model in generating predicted labels for the model inputs in the candidate batch of model inputs, and
(ii) a diversity of the model inputs in the candidate batch of model inputs;
wherein for each candidate batch of model inputs, generating the score for the candidate batch of model inputs comprises:
identifying a plurality of pairs of model inputs that each include a respective first model input and a respective second model input from the candidate batch of model inputs;
determining, for each pair of model inputs in the candidate batch of model inputs, a respective covariance between: (i) a predicted label for a first model input in the pair of model inputs, and (ii) a predicted label for a second model input in the pair of model inputs; and
generating the score for the candidate batch of model inputs based on the respective covariance for each pair of model inputs in the candidate batch of model inputs; and
selecting the current batch of model inputs from the set of candidate batches of model inputs based on the scores;
obtaining a respective target label for each model input in the current batch of model inputs, wherein a target label for a model input defines a model output that should be generated by the machine learning model by processing the model input; and
training the machine learning model on at least the current batch of model inputs using the target labels for the current batch of model inputs; and
outputting the trained machine learning model.