Anomaly Detection using Variational Autoencoder with Spectrum Analysis for Time Series Data

Uncertainty is an ever present challenge in life. To meet this challenge in data analysis, we propose a method for detecting anomalies in data. This method, based in part on Variational Autoencoder, identifies spiking raw data by means of spectrum analysis. Time series data are examined in the frequency domain to enhance the detection of anomalies. In this paper, we have used the standard data sets to validate the proposed method. Experimental results show that the comparison of the frequency domain with the original data for anomaly detection can improve validity and accuracy on all criteria. Therefore, analysis of time series data by combining Variational Autoencoder and frequency domain spectrum methods can effectively detect anomalies. Contribution- We have proposed an anomaly detection method based on the time series data analysis by combining Variational Autoencoder and Spectrum analysis, and have benchmarked the method with reference to recent related research.