A Survey on Time Series Online Sequential Learning Algorithms
Rahul Katarya, T. Sai Lalith Prasad · 2017
Time-series analysis is extremely important in various fields. In this survey paper, an overview of existing Online Sequential Learning Algorithms for time series modeling and analysis will be provided. Because of their good capability of approximation, researchers are looking to work forward with Artificial Neural Networks. Time series is a very important part of data sets which are temporal in nature and they are found abundantly in various applications. We provide different existing Online Sequential algorithms and for each category, we identify the advantages and disadvantages of the techniques in that category. We have done a survey on the time series algorithms by including various recent papers. In this survey paper, we are also going to show the Multiple Model Switching and Tuning approach (MMST) description since it increases the accuracy of the model.