METHODS AND TOOLS FOR HYBRID INFERENCE IN FILE MANAGEMENT IN ARTIFICIAL INTELLIGENCE / MACHINE LEARNING SYSTEMS

Abstract

The WTRU is configured with input and output file patterns and AI / ML models with trained weights. Each input file pattern includes file indices for a set of input files, and each output file pattern includes file indices for a set of output files. Each AI / ML model corresponds to one output file pattern. The WTRU measures the signal quality of a subset of the input file set of each input file pattern to identify the best input file and reports it to the base station. The WTRU receives, from the base station, indications about one of the input file patterns and one of the output file patterns to predict the best output file based on the reports.WTRU predicts the best output beam using an AI / ML model corresponding to the indicated output beam pattern and trained weights and reports the predicted best beam to the base station.