5. (Classification 10 pt in total; breakdown for the individual parts given below) Suppose you were constructing a Naive Bayes classifier to classify if a new patient entering the hospital and exhibiting lung disease has Bronchitis or Tuberculosis. Suppose you take a sample of 6 patients and suppose your data set looks like the following table. Figure 1: Training data (a) (2 pt) What are the features you are using to make this classification? (b) (3 pt) If you are using Naive Bayes classifier, what are you assuming about these features? 4 (c) (5 pt) Let us continue to assume you will be using Naive Bayes for classification. Suppose a new patient comes in with symptoms as follows: Using the training data to construct a Naive Figure 2: New patient Bayes classifier, would you classify this patient as Tuberculosis or Bronchitisnusing Naive Bayes estimator?

5. (Classification 10 pt in total; breakdown for the individual parts given below) (you might be able to do this only after Thursday's class) Suppose you were constructing a Naive Bayes classifier to classify if a new patient entering the hospital and exhibiting lung disease has Bronchitis or Tuberculosis. Suppose you take a sample of 6 patients and suppose your data set looks like the following table. Disease X-ray Shadow Dyspnea Lung inflammation Bronchitis Yes Yes Yes Bronchitis Yes Yes Yes Bronchitis No No Yes Tuberculosis No No Yes Tuberculosis Yes Yes No Tuberculosis Yes No No Figure 1: Training data (a) (2 pt) What are the features you are using to make this classification? (b) (3 pt) If you are using Naive Bayes classifier, what are you assuming about these features?(c) (5 pt) Let us continue to assume you will be using Naive Bayes for classification. Suppose a new patient comes in with symptoms as follows: Using the training data to construct a Naive X-ray Shadow Dyspnea Lung inflammation Yes No Yes Figure 2: New patient Bayes classifier, would you classify this patient as Tuberculosis or Bronchitisnusing Naive Bayes estimator