Analysis of time series data for anomaly detection

Abstract

The integration of sensors in our everyday lives and in industry presents a serious challenge to data analysis professionals. Since the use of smart devices has exponentially increased the amount of data collected in all areas, we must not only store these data, but also extract valuable information from those using some data analysis method. In many cases, the increased amount of data also causes problems for data analysis algorithms, so we need to be continuously updated and specialized for fundamental purposes. In the article, we will present some data analysis techniques, starting from the simplest statistical methods to more complex techniques using machine learning and presenting one of their possible applications. In our study, we will use the KMeans clustering algorithm and examine its effectiveness in time series sensor data analysis.