Forecasting Outlier Occurrence in Stock Market Time Series Based on Wavelet Transform and Adaptive ELM Algorithm

Abstract

In financial field, outliers represent volatility of stock market, which plays an important role in management, portfolio selection and derivative pricing. Therefore, forecasting outliers of stock market is of the great importance in theory and application. In this paper, the problem of predicting outliers based on adaptive ensemble models of Extreme Learning Machines (ELMs) is considered. We found out that the proposed model is applicable for outlier forecasting and outperforms the methods based on autoregression (AR) and extreme learning machine (ELM) models.