Question

1.This is a popular way to measure impurity. A.Gini Index B. Entrophy C. Both a and b D. None of the above

2.The numbers on the left fork at a decision node are the number of records in the decision node that had A.Values equal to splitting value B.Values greater than the splitting value C.Values less than or equal to splitting value D.None of the above

3.Using a Full grown tree (based on training data) leads to A.Outliers B.Misclassification errors C.Complete overfitting of data D.None of the above

4.Overfitting can be limited by A.Stopping Tree Growth B.Pruning the full grown tree C.Both a and b D.None of the above

5.This is a recursive partitioning method that predates classification and regression tree (CART) A.Pruning B.Association C.CHAID (chi-squared automatic interaction detection) D.None of the above

6.Different cutoff values lead to different classifications and consequently different classification matrices. The overall accuracy is computed for various values of the cutoff value, and the cutoff value that yields maximum accuracy is chosen. The danger is A.Outliers B.Misclassification C.False positives D.Overfitting

7.In logistic regression, the relation between Y and Beta parameters is non linear. For this reason, Beta parameters are estimated using A.Maximum likelihood method B.Method of least squares C.Both a and b D.None of the above

8.Logistic Regression can be used for A.Classification B.Profiling C.Both Classification and Profiling D.None of the above

9.Logistic Regression can be used in data to find similarities between observations within each class in terms of the predictor variables. This is known as A.Probability of variables B.Classification C.Profiling D.All the above

10.This Metric is very popular in sports, gambling in general, epidemiology, and many other areas and is defined as the ratio of the probability of belonging to class 1 to probability to belonging to class 0 A.Probability of winning B.Probability of contracting a disease C.Classification D.Odds

Answer 1

1) The popular way to measure impurity is a) Gini Index

2) The numbers on the left fork at a decision node are the number of records in the decision node that had c) Values less than or equal to splitting value