Title: Prediction of financial crises using statistic model and intelligent technologies in ubiquitous environments
Authors: Junsuke Senoguchi; Setsuya Kurahashi
Addresses: Graduate School of Business Science, University of Tsukuba, 3-29-1 Ootsuka, Bunkyo-ku, Tokyo, 112-0012, Japan ' Graduate School of Business Science, University of Tsukuba, 3-29-1 Ootsuka, Bunkyo-ku, Tokyo, 112-0012, Japan
Abstract: A large number of earlier studies partially revealed the mechanism of the financial crises in recent years. However, no study has yet conducted the variable selection from the mounds of factors including the critical component of key macro financial statistics. This study, using a traditional logistic model and intelligent system technologies, explored some key influential factors on the occurrence of the financial crises. As a result, the cyclical component of the current account as percentages of GDP and the cyclical component of the domestic loan as percentages of GDP have been proven a key factor to predict a financial crisis. In the present, China and Malaysia can be classified as a crises group with 77% of possibilities.
Keywords: intelligent technologies; classification; regression tree; CART; variable selection; artificial intelligence; financial crisis prediction; financial crises; statistic models; ubiquitous environments; logistic models; modelling; current accounts; GDP; gross domestic product; cyclical components; domestic loans.
International Journal of Computer Applications in Technology, 2013 Vol.48 No.2, pp.173 - 183
Received: 08 May 2021
Accepted: 12 May 2021
Published online: 24 Aug 2013 *