Gene Chip Data Analysis of Support Vector Machines with Different Kernel Function

• Main Authors: LungHsin-Wang, 王瀧興
• Other Authors: Shyh-chang Tsaur
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/32565765341662675037

碩士 === 亞洲大學 === 資訊工程學系碩士班 === 98 === Biological gene chip has been covered by a layer veil of mystery. Great impact in human and biological. According to the Department of Health statistics of women suffering from ovarian cancer incidence and mortality are the top ten in Taiwan. In particular, case fatality rate the top. Incidence is also second only to cervical cancer. In this paper, the ovarian cancer with mouse training database analyzed for solve many variables and sample size small of gene chip. For this reason, linear regression and analysis of variance (ANOVA) are used to pre-processing done to reduce dimensional and identify valuable gene. Furthermore, the information database to divided and examined by support vector machine (SVM) with comparison of the kernel function experimental results. We to discover the classification results are good and different kernel function is not same as the performance of SVM. Finally, the results of the discussion and to identify the most efficient high accuracy rate of the Kernel Function.