LSTM Autoencoder for anomaly detection

LSTM Autoencoder for anomaly detection in ECG time series. I developed an adaptive threshold selection strategy to identify anomalies on the basis of mean reconstruction error obtained from training data.