Anomaly Detection using LSTM Autoencoders

Anomalies in time series data can indicate unusual or unexpected events, such as a sudden spike or drop in stock prices, an extreme weather event, or a sudden increase in website traffic. Deep learning techniques such as recurrent neural networks (RNNs) , long short term memory (LSTM) etc. can be used to model the temporal dependencies in time series data. In this project we have used LSTM Autoencoder to detect anomalies based on the analysis of time-series data. We have applied the model to three different datasets with slight variations as needed. The AI deep learning neural network used in this project for anomaly detection is built using Python, Keras and TensorFlow.