FIELD: machine learning.
SUBSTANCE: invention relates to a method and system for controlling an electronic device using a machine learning model. The system includes a processor for receiving training data for a user activity; accepting rejection criteria; defining a set of model parameters for a machine learning model, which includes: (1) applying the machine learning model to the training data; (2) generation of forecast errors based on the model; (3) generating a data selection vector to identify non-outlier target variables based on model-based prediction errors; (4) using a data selection vector to form a non-outlier data set; (5) determining updated model parameters based on the non-outlier data set; and (6) repeating steps (1)-(5) until the censoring completion criterion is satisfied; training classification model parameters for a machine learning model of outlier value classifiers; applying a machine learning model of outlier value classifiers to activity-related data to identify non-outlier activity-related data; applying a machine learning model to non-outlier activity-related data to predict future activity-related attributes for a user activity; and generating a machine control command based on applying the machine learning model and predicting the future of the activity-related attribute associated with the user activity.
EFFECT: improved accuracy of predicting user activity attributes for generating a machine command to control an electronic device.
20 cl, 16 dwg, 12 tbl
