臺灣博碩士論文加值系統

「簡單貝氏分類法」是一種透過各特徵彼此間互相獨立下運用貝氏定理的分類器，現實生活中大部分的資料無法滿足任何兩個特徵皆不相關這項假設。
我們提出一種結合主成分分析與費雪資訊之簡單貝氏分類法，先利用「主成分分析」將各特徵轉化成任何兩個特徵皆不相關的新的特徵，並將各屬性欄位與類別欄位進行費雪資訊的計算，逐一選出具有較多資訊量的屬性，最後估計貝氏分類法所需的相似度，並將物件歸類於最高「事後機率」的類別。
我們以網路常用的標準資料集為研究與效能評估，探討利用主成分分析和費雪資訊降低特徵維度，此分類法準確率的變化。

Naive Bayes classifier is a simple probabilistic classifier which is based on applying Bayes’theorem which strong independence assumptions between the features.
We propose a method based on Naive Bayes classifier with Principal Components Analysis(PCA) and Fisher Information.
We use Principal Components Analysis to make features uncorrelated.
The transformed features are ranked by Fisher Information score which measuring the amount of information and calculate the posterior probability where the likelihood is replaced by p-value.
We conclude our research through the classification accuracy with some examples and present our vision for future research.

摘要
Abstract
致謝
目錄
圖目錄
表目錄
1 研究背景
1.1 簡單貝式分類法 (Bayes classifier)
1.2 主成分分析 (Principal Components Analysis)
1.3 費雪資訊 (Fisher Information)
2 研究方法
2.1 利用 p 值 (p-value) 進行分類
2.2 ROC 曲線下面積 (AUC)
2.3 R 平方
2.4 費雪資訊
3 實驗
3.1 資料集說明
3.1.1 Wisconsin Diagnostic Breast Cancer(WDBC)
3.1.2 Diabetic Retinopathy Debrecen(DRD)
3.2 實驗步驟
4 結論
4.1 Wisconsin Diagnostic Breast Cancer(WDBC)
4.2 Diabetic Retinopathy Debrecen(DRD)
4.3 結論
參考文獻

Hsu, C. N., Huang, H. J., and Wong, T. T.. “Implications of the Dirichlet assumption for discretization of continuous attributes in naïve Bayesian classifiers”. Machine Learning, 53, 235-263, 2003.

Mohammad Darzi, Ali AsgharLiaei, Mahdi Hosseini, HabibollahAsghari. "Feature Selection for Breast Cancer Diagnosis: A Case-Based Wrapper Approach. " (2011), World academy of Science, Engineering and Technology 77, 2011,pp 1142-1143.

E.Osuna, R.Freund, and F. Girosi, “Training support vector machines: Application to face detection”.
Proceedings of computer vision and pattern recognition, Puerto Rico pp. 130–136.1997.

D. Lavanya, “Ensemble Decision Tree Classifier for Breast Cancer Data,” International Journal of Information Technology Convergence and Services, vol. 2, no. 1, pp. 17-24, Feb. 2012.

Ch. Rakesh, , D.N.D.Harini, , M. Bhanu Sridhar "An Empirical Analysis of Classification Algorithms for Medical Data"

Lichman, M. (2013). UCI Machine Learning Repository [http://archive.ics.uci.edu/ml]. Irvine, CA: University of California, School of Information and Computer Science.

Balint Antal, Andras Hajdu: An ensemble-based system for automatic screening of diabetic retinopathy, Knowledge-Based Systems 60 (April 2014), 20-27.

Frieden, B. R. (2004) Science from Fisher Information: A Unification. Cambridge Univ. Press. ISBN 0-521-00911-1.