Consider the traffic accident data set shown in the table.

Let Crash Severity be the class label. Show the key steps for the following tasks. Note: Task (b) is 10-point extra credit.

(a) Using information gain as the attribute selection measure, construct the first level of the decision tree.

(b) If gain ratio is used as the attribute selection measure, will the first level of the decision tree be different from above?

(c) Given a traffic accident with the values Good, Sober, None, and Yes for the attributes Weather Condition, Drivers Condition, Traffic Violation, and Seat Belt, respectively, how would a nave Bayesian classifier determine whether Crash Severity would be Minor or Major?

Consider the traffic accident data set shown in the tolowing table. Seat Crash Belt Severity Driver's Condition Sober Sober Sober Traffic Violation None Disobey stop sign Disobey stop sign Weather Condition Bad Good Bad Good Alcohol-impaired Exceed speed limit NoMajor Bad Good Alcohol-impaired Disobey stop sign Yes Minor Good Alcohol-impaired Exceed speed limit Yes Major Bad Good Good Alcohol-impaired Bad Good nor Yes Minor Yes Minor es Sober Disobey traffic signal NoMajor Yes Major Disobey traffic signal YesMajor No Major Disobey traffic signal NoMajor Exceed speed limit Yes Major Alcohol-impaired None Sober None Sober Sober Let Crash Severity be the class label. Show the key steps for the following tasks. Note: Task (b) is 10-point extra credit (a) Using information gain as the attribute selection mensure, construct the first level of the decision tree (b) If gain racio is used as the attribute selection measure, will the first level of the decision tree be different from above? (c) Given a trattic accident with the values "Good". Sober", "None", and "Yes" for the attributes Weather Condtion Driver's Condition Tramc Violation, and Seat Beit, respecively, how would a naive Bayesian classifier determine whether Crash Severity would be Minor or Major