Title

Hybrid Approach: Neural Networks and Japanese Candlestick Method for Financial Forecast

Abstract

This paper shows the experimental study of the cnadlestick method in the hybrid financial forecasting models. A committee machine with the Generalized Regression Neural Netowrok (GRNN) experts is the primary tool that handles the input data, and the candlestick method is introduced to the model using gating networks. This introduction of the candlestick method into stock quotes of Exxon Moblie, General Electrics, General Motors, Google, Microsoft, and Wells Fargo are used as input data sets. The output of the model is the forecast of the next day's closing proce. For the purpose of comparison, the performance of a sinple GRNN-based forecasting model is shown. The results of their forecasts are evaluated on the basis of the mean squared error.

Department(s)

Engineering Management and Systems Engineering

Keywords and Phrases

Candlestick Method; Data Sets; Generalized Regression Neural Network (GRNN); Hybird Financial Forcasting Models

Document Type

Article - Conference proceedings

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 2006 American Society of Mechanical Engineers (ASME), All rights reserved.

This document is currently not available here.


Share

 
COinS