Applying Cloud Computing to the Identification of Students with Learning Disabilities – Using Amazon Elastic Compute Cloud as an Example

• Main Author: 吳宗修
• Other Authors: 吳東光
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/36204215129119498005

碩士 === 國立彰化師範大學 === 資訊管理學系所 === 99 === Learning disability (LD) is a kind of internalizing disorders, and is difficult to identify from looks. The identification of LDs is very complicated and may take a lot of time and resources. Previous studies had used the algorithm like Parallel Distributed Genetic Algorithms combining with Artificial Neural Network to develop a system that assists the identification of LDs. The result of these researches showed a satisfactory correct identification rate (CIR). In recent years, the efficiency of computer hardware has improved, but most applications are unable to fully utilize the efficiency. Fortunately, the virtualization technology allows virtual machines to make full use of the hardware resources, and improves the hardware efficiency accordingly. With the growing popularity of the mobile internet services, users can use the resources of internet by mobile internet devices. We try to apply the computing power of cloud computing to analyze the identification data of LDs. By using more computing nodes and more computing generations, we have improved the CIR. Our experiments shows that while fixing the total population sizes and using more computing nodes to reduce the populations processed by each node, the computation time and the accuracy in LDs diagnosis can be both improved. To avoid local optimum, the subpopulation size is under consideration.