HW7

Dear all,

Just “one” question this time. :)

1.

a. Try to build a decision tree (not using library tho) to predict if someone will play golf or not based on this toy dataset.

For simplicity, let's ignore the numerical value for humidity and only use the categorical column (high/normal). And for the temperature value, let's make it categorical also and only consider if the temperature is \(>\) average or \(\le\) average.

b. Let's try to use the temperature value more efficiently. Instead of asking a question of if temperature is \(>\) average or \(\le\) average, let's pick an optimal threshold T and consider temperate is \(>\) T or \(\le\) T instead. Will your decision tree change after this modification?

Best,

Sam