Deep Multiscale Soft-Threshold Support Vector Data Description for Enhanced Heavy-Duty Gas Turbine Generator Sets’ Anomaly Detection

Table 2

The backbone of the AE network.

---

Algorithm 1: The algorithmic procedure for the first phase

Input: The training set , batch size B, the learning rate a, iteration count K

Output: Encoder ψ(·), decoder φ(·)

Initialization: Encoder ψ(·), decoder φ(·), parameter set and

(1) for k = 1 to K do:

(2)  Randomly select a mini-batch of data from the training set

(3)  Obtain the reconstructed samples from the training samples

(4)  

(5)  ,

(6) end for

---