Selecting Feature Types to Extract Based on Pre-Classification of Sensor Measurements

Abstract

A processing apparatus including one or more processors and memory receives sensor measurements generated by one or more sensors of one or more devices, pre-classifies the sensor measurements as belonging to one of a plurality of pre-classifications, and selects one or more feature types to extract from the sensor measurements based at least in part on the pre-classification of the sensor measurements. The processing apparatus also extracts features of the one or more selected feature types from the sensor measurements and determines a state of a respective device of the one or more devices in accordance with a classification of the sensor measurements determined based on the one or more features extracted from the sensor measurements.

What is claimed is: 
     
         1 . A method comprising:
 at a processing apparatus having one or more processors and memory storing one or more programs that, when executed by the one or more processors, cause the respective processing apparatus to perform the method:
 receiving sensor measurements generated by one or more sensors of one or more devices; 
 pre-classifying the sensor measurements as belonging to one of a plurality of pre-classifications; 
 selecting one or more feature types to extract from the sensor measurements based at least in part on the pre-classification of the sensor measurements; 
 extracting features of the one or more selected feature types from the sensor measurements; and 
 determining a state of a respective device of the one or more devices in accordance with a classification of the sensor measurements determined based on the one or more features extracted from the sensor measurements.