An LSTM-Autoencoder Architecture for Anomaly Detection Applied on Compressors Audio Data

Table 10

Model dimensions’ overview and operations for LSTM employing SoftMax layer (created by authors).

Operation		Data	Dimensions	Weights (N)	Weights (%)
Input	#####	2	259
LSTM	LLLLL	—	—	62,000	99.8%
tanh	#####		50
Dropout	|||	—	—	0	0.0%
	#####		50
Dense	XXXXX	—	—	102	0.2%
SoftMax	#####		2