COMP4388: Machine Learning

Please answer the following questions with each in no more than 2 pages. Please make sure to include proper references and citations. Q1) Bagging and Boosting are two methods for ensemble learning in machine learning. Please answer the following: a. What is ensemble learning? b. What is bagging? c. What is boosting? d. Explain in your own words the difference between bagging and boosting. Q2) In this part, you will experiment the difference in the behavior of SVM kernels. In your own words, explain how these kernels (i.e., linear, polynomial of 2nd and 3rd degrees, and RBF) differ in means of accuracy on the generated models. You should perform the experiment using the Weka1 data mining tool and use the binary class dataset found under this link(https://www.dropbox.com/s/8r3nfghlhg96n63/COMP4388- Dataset1.csv?dl=0). Split the data 80 training/20 test and explain the steps of your work (i.e., if you scale data, how did you split, provide appropriate figures/plots of the data, confusion matrix,). Show the difference in Accuracy and F-score between kernels.