Sequential ELM for financial markets

This paper deals with time-series prediction using artificial neural networks in the context of financial markets. Specifically, in this paper we consider the prediction of the Oil & Gas Index of the Bombay Stock Exchange. Two classes of training strategies are compared in this paper. The first class is based on the back propagation algorithm and the second class is based on the extreme learning machine. The primary objective is to demonstrate that the prediction performance of the recently proposed sequential variant of the extreme learning machine is superior to other training strategies considered here with the added advantage of lesser computation time. For the back propagation algorithm, the paper also proposes combining batch and online training phases to enhance predictive performance.