Example Name Will buy A Positive B Negative C Negative D Positive E Negative FA Positive Positive Negative Positive Nega

[answerhappygod]

Example Name Will Buy A Positive B Negative C Negative D Positive E Negative Fa Positive Positive Negative Positive Nega 1 (93.13 KiB) Viewed 11 times

Example Name Will buy A Positive B Negative C Negative D Positive E Negative FA Positive Positive Negative Positive Negative G He Class attribute Je Age >40< 25..40 25..40 >40< 25..40€ >40< 25..40 <25< 25..40 <25< Feature attributes Education Good Poor Fair Good Good Fair Fair Poor Good Good Credit_rating Excellent Good Good Good Good Excellent Good Good Good Good
Suppose the feature attribute Age will be selected as the first splitting node in decision tree. Consider the branch Age = 25..40 in Figure 1, find the information gains of the feature attributes Education and Credit rating. Table 6 can be used in performing the calculation. (10 marks) Age? <25 >40 Class: Negative Prob. 2/2 ? 25...40 Class: Positive Prob. 3/3 Find the information gains of the features Education and Credit_rating (
Gain Education C1 (good) C2 (Fair) C3 (poor) Gain Credit_rating C1 (Excellent) C2 (Good) 0.4 0.6 0.9710 - Parent Entropy Positive Negative Sub-Total Positive (rate) Negative (rate) <- Total 0.4 0.6 0.9710 - Parent Entropy Positive Negative Sub-Total Positive (rate) Negative (rate) <- Total Entropy(ci) p(ci) Entropy(ci) p(ci) Weighted Sum of Entropy <- Information Gain Weighted Sum of Entropy <- Information Gain