An Anomaly Detection Algorithm Selection Service for IoT Stream Data Based on Tsfresh Tool and Genetic Algorithm

Table 4

Description of our experimental datasets.

Model	Hyperparameter	Value
LOF	Distance function (k)	10, Minkowski distance
PCI		30, 98.5
LSTM	Filters, optimisers, architecture, loss, batch size, and epochs	4 ∗ 4, Adam, 2-state full LSTM layer, MSE, 32, 50
OC-SVM	Upper bound of outliers, kernel	Radial basis function kernel (RBF), 0.7
Autoencoder	Architecture, activation functions, optimiser, loss, batch size, and epochs	Decoding layers (16, 32), encoding layers (32, 16), linear for output, ReLU for encoding and decoding, MSE, Adam, 32, 50