HYBRID SWARM INTELLIGENCE-BASED SYSTEM TO EVALUATE FORECASTING APPLICATIONS IN CIVIL ENGINEERING AND MANAGEMENT

• Main Author: Anh-Duc Pham
• Other Authors: Jui-Sheng Chou
• Format: Others
• Language: en_US
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/qh9wh5

博士 === 國立臺灣科技大學 === 營建工程系 === 102 === Advanced data mining (DM) techniques are potential tools for solving civil engineering and management (CEM) problems.This study investigated the potential use of various advanced approaches and proposes a novelsmart artificial firefly colony algorithm (SAFCA)-based support vector regression models (SAFCAS) that integrates firefly algorithm (FA), chaotic maps, adaptive inertia weight, Levy flights, and support vector machine-based regression (SVR). Firstly, adaptive approach and randomization methods are incorporated in FA to construct a novel and highly effective meta-heuristic algorithm for global optimization. The enhanced FA is then used to identify the optimal set of turning parameters in SVR model. The proposed system is validated by comparing the performance of the SAFCAS with those of empirical methods, well-known AI models, and previous works via cross-validation algorithm and hypothesis test.For real-world engineering cases, eight datasets are collected from reliable laboratories and published literature. Experimental results obtained from theSAFCASconfirm that using the proposed hybrid system in advanced DM approaches significantly improve the accuracy of forecasting methods used to solve real-life CEM problems.