Machine learning approaches for time series analysis

Abstract

Time series data can be obtained from everywhere in everyday life It is constantly newlinegenerated from various human activities and different real world applications such as newlinebiomedical signals weather recordings stock exchange rates and many more Generally time series data can be categorized into two types namely univariate time series and multivariate time series The analysis of time series gave birth to different tasks namely indexing forecasting clustering and classification In this thesis particularly we deal with 3 time series problems which are representation classification and forecasting