4. The following table lists a dataset from the credit scoring domain we discussed in the chapter. Underneath the table we list two prediction models that are consistent with this dataset, Model 1 and Model 2.

a. Which of these two models do you think will generalise better to instances not contained in the dataset?

b. Propose an inductive bias that would enable a machine learning algorithm to make the same preference choice as you did in part (a).

c. Do you think that the model that you rejected in part

(a) of this question is overfitting or underfitting the data?

Transcribed Image Text:

ID OCCUPATION AGE LOAN-SALARY RATIO OUTCOME 1 industrial 39 3.40 default 2 industrial 22 4.02 default 3 professional 30 2.70 repay 4 professional 27 3.32 default 5 professional 40 2.04 repay 6 professional 50 6.95 default 7 industrial 27 3.00 repay 8 industrial 33 2.60 repay 9 industrial 30 4.50 default 10 professional 45 2.78 repay Model 1 if LOAN-SALARY RATIO > 3.00 then OUTCOME=default else OUTCOME = repay Model 2 if AGE= 50 then