Federal learning 2.1 created by Xiao Peng, Lin Ze, Hao Ying, Dr. Wang Chuang
fault diagnosis.py is the main processing file for "overall model" xiao_fault diagnosis.py is the main processing file for "vote model"
model.py define several CNN model, which "CNN2d_classifier_xiao" is used for clip size=2304; "CNN2d_classifier" is used for clip size=4096 (only for previous data precessing methods); "CNN2d_fitting_xiao" is used for clip size=576; other model is not used yet.
xiao_estimate.py is used for vote model estimate, it load each part-model and test accuracy of vote model.
xiaodataset.py is for data process, compare to the previous methods, it can increase the utilization rate of data. xiao_dataset_random.py is for data process too, this can divide the data into train and test randomly.