臺灣大學: 數學研究所鄭明燕林筱淋Lin, Hsiao-LinHsiao-LinLin2013-03-212018-06-282013-03-212018-06-282011http://ntur.lib.ntu.edu.tw//handle/246246/249825在非參數判別分析中，我們利用區間Logistic迴歸模型 (Local logistic regression)估計貝氏準則的事後機率。在進行區間Logistic迴歸時，我們需要決定平滑參數值，我們取使得GCV值最小的平滑參數，之後再進行區間Logistic迴歸，我們的摸擬實驗發現這樣會得到理想的分類正確機率。In nonparametric discriminant analysis, we use local logistic regression model to estimate the posterior probability of Bayes rule. Before we carry on local logistic regression, we need to choose the smoothing parameter. We choose the smoothing parameter by minimizing the GCV. A simulation study suggests that this approach performs well.170462 bytesapplication/pdfen-US非參數判別分析分類Logistic 迴歸貝氏準則nonparametricdiscriminant analysisclassificationLogistic regressionBayes rulethesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/249825/1/ntu-100-R95221028-1.pdf