Application of the Hadoop-based Parallel Particle Swarm Optimization to the Identification of Students with Learning Disabilies

• Main Authors: Yu-Jie Ciou, 邱俞潔
• Other Authors: Tung-Kuang Wu
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/84095809379015213740

碩士 === 國立彰化師範大學 === 資訊管理學系所 === 100 === The process in identification of student with learning disabilities (LD) is complicated, time-consuming and requiring extensive manpower and resource. Previous researches used Artificial Neural Network (ANN) to assist the diagnosis of students with learning disabilities and showed pretty good performance. However, the construction of ANN-based classification model through genetic algorithm may take pretty much time. Accordingly, grid and cloud computing has been used to speed up the process. Instead of the genetic algorithm, this study uses particle swarm optimization (PSO) algorithm in constructing the ANN-based LD identification model. Hadoop-based cloud computing environment is also used in this study to assist the optimization process. The experimental results show that PSO-based algorithm may achieve better correct identification rate, while take a little longer time, as compared to the genetic-based algorithm.