An Anomaly Detection Approach Based on Integrated LSTM for IoT Big Data

Algorithm 1

Anomaly Detection Algorithm Based on C²-LSTM.

	Input: are the input sets, and label is the corresponding
	Output:A trained anomaly detection model M
(1)	Initialize the model M
(2)	Initialize the iteration count T, batch size N, threshold δ
(3)	for q = 1 t o T do
(4)	for m = 1 to 2 do
(5)	for each batch do
(6)	Transfer into via CNN by equation (7)
(7)	Transfer into via CNN by equation (8)
(8)	Splice and into S
(9)	Predict based on Z via the estimation network
(10)	Update M to minimize loss
(11)	end for
(12)	if loss < δ: break
(13)	end for
(14)	return M