Model based Machine Learning

The training data consists of data from various faults from five individual configurations, while the testing data is blind and is from one individual configuration of the rock drill.  A final validation data set will be from two individual configurations from the rock drill and the labels are blind. 

The training data set contains data from 11 different fault classification categories, in which 10 are different failure modes and one class is from the healthy/no fault condition. 

Each file follows the naming convention of data_{sensor}{individual impact cycle number}.csv.