A study on neural networks approach to time-series analysis

Prediction and classification of time-series data play a vital role in multiple fields. This paper aims to compare modern machine learning techniques to analyze time series data. A discussion of existing algorithms is presented. Artificial Neural Networks have seen major applications in various computational fields in diverse ways. Its application extends to time-series data in an essential way. A number of recently developed Artificial Neural Network approaches to time-series prediction are provided. We discuss merits and demerits of each algorithm are. We then provide a comparative study of the approaches and address their applications.