Research Article
LstFcFedLear: A LSTM-FC with Vertical Federated Learning Network for Fault Prediction
Algorithm 1
LSTM-FC network algorithm.
INPUT PARAMETERS: Reasonable labeling T+,T− | Label labeling is not reasonable ˜ T+, ˜ T− | Maximum number of pre-training | Maximum number of training | OUTPUT VALUE: Train well-performing models | 1: d, c ←0 | 2: get x on T | 3: while d < preEpoch do | 4: for each fault i in T do | 5: get ˜r+ i ,˜r−i in L | 6: get L by (9) on i; | 7: update θ; | 8: compute Lf | 9: update θ; | 10: end for 11 j ← j +1; 12 end while | 13: while k < trainEpoch do | 15: for each fault xi in T do | 16: compute xi; | 17: compute yi; | 18: compute LA; | 19: update θ; | 20: end for | 21: k ← k +1; | 22: end while |
|