Description
2. Repeat the experiment using Naive Bayes Classifier.
3. Use the Naive Bayes and K-NN classifiers on the Breast Cancer (Wisconson) dataset. Again the data has been downloaded as breast.csv in moodle. Note that the Breast Cancer dataset contains 11 attributes. The first one is the patient id, which can be ignored. The last one is the class label, 2 for negative (benign) and 4 for positivbe (malignant). The rest features are features valued from 1 to 10. There is a description of all features in the official site. Note that some features are imported as string values, (in particular, feature 6), and has to be converted to numerical data first.
Repeat the K-NN classification with k = 1,3,5,7.
For each of the classification methods, since the data file is shuffled before splitting between the training and test set, repeat the experiments 10 times and find the average accuray. Write a short report and discuss your result.
1
