SVM

To predict the wine quality using the support vector machine classification algorithm.

(a) Download the wine quality data and use the training dataset to conduct the grid-search to find the optimal hyperparameters of svm by using the linear kernal.

(b) Train a svm classifier by using the linear kernal and the corresponding optimal hyperparameters, then make predictions on the testing dataset, report the predictive performance.

(c) Conduct the grid-search to find the optimal hyperparameters of svm by using the RBF kernal.

(d) Train a svm classifier by using the RBF kernal and the corresponding optimal hyperparameters, then make predictions on the testing dataset, report the predictive performance.

(e) Conduct the ROC curve analysis to compare the predictive performance of svm classifiers trained by using the linear and RBF kernels respectively.

Hint: Given a pre-defined hyperparameter space - C: [0.01, 0.1, 1, 5, 10], and : [0.01, 0.03, 0.1, 0.5, 1].

WineQuality Training data

fixed acidity,volatile acidity,citric acid,residual sugar,chlorides,free sulfur dioxide,total sulfur dioxide,density,pH,sulphates,alcohol,quality 9.1,0.27,0.45,10.6,0.035,28,124,0.997,3.2,0.46,10.4,Good 6.6,0.36,0.29,1.6,0.021,24,85,0.98965,3.41,0.61,12.4,Good 7.4,0.24,0.36,2,0.031,27,139,0.99055,3.28,0.48,12.5,Good 6.9,0.36,0.34,4.2,0.018,57,119,0.9898,3.28,0.36,12.7,Good 7.1,0.26,0.49,2.2,0.032,31,113,0.9903,3.37,0.42,12.9,Good 6.2,0.66,0.48,1.2,0.029,29,75,0.9892,3.33,0.39,12.8,Good

WineQuality Testing data

fixed acidity,volatile acidity,citric acid,residual sugar,chlorides,free sulfur dioxide,total sulfur dioxide,density,pH,sulphates,alcohol,quality 4.4,0.46,0.1,2.8,0.024,31,111,0.98816,3.48,0.34,13.1,Good 6.6,0.38,0.29,2.9,0.035,15,101,0.98916,3.04,0.37,12.5,Good 6.4,0.28,0.22,12.8,0.039,51,150,0.99535,3.23,0.43,10.7,Good 6.9,0.29,0.25,12.2,0.04,29,136,0.99552,3.05,0.65,10.4,Good 6.6,0.38,0.29,2.9,0.035,15,101,0.98916,3.04,0.37,12.5,Good 6.6,0.18,0.26,17.3,0.051,17,149,0.9984,3,0.43,9.4,Good