A framework for neural network to make business forecasting with hybrid VAR and GA components

Applying Vector Autoregression (VAR) and genetic algorithm (GA) in hybrid systems with neural network can improve the NN's prediction capability. Two case studies have been carried out to demonstrate how to build our VAR-NN-GA system and its advantages. One is on the tourist patterns. Relatively recently, neural network is introduced into the tourist forecasting field [1, 2, 3]. Results show that the hybrid forecasting system is more robust and able to select variables automatically and makes more accurate prediction than the stand-alone neural network. Another case study is the Asian Pacific stock markets, which is more complicated. There exist strong time-dependent correlations between the US market and the Asian markets. While it is difficult to model the markets' interactions with a static model, VAR and GA are adopted for the dynamic model-selection process. The results for the stock market case study show that the hybrid system is about 30% better than the best individu...

• Type: Journal
• Year: 2006
• Where: ENGL
• Authors: Sio Iong Ao
• Keywords: Case Study, ENGL 2006, Hybrid System, Neural Network