models Pros and cons

Question 1

GLM vs trees

Answer 1

Trees do not require assumption about data.
Trees are easy to interpret and explain
Trees are easy to present visually
Trees are easy to construct
Trees dont use dummy variables for factor
Interactions in tree are automatically handled
Trees remove insignificant predictors automatically

Disadvantage of trees
Risk of overfitting
Not robust, could vastly change using a subset (sensitive to training data)
Recursive binay is greedy