Tentamen Data Mining Augustus 2005

Data Mining Exam, August 2005

Questions can be answered in Dutch or English.

1. Explain what the “language bias” and what the “search bias” are in a machine learning system.

2. What are association rules?

3. What is instance based learning?

4. What is stratified threefold cross-validation?

5. What is a Lift chart?

6. How can "decision tree" algorithms deal with numerical values?

7. Explain what pre-pruning and post-pruning are in the construction of decision trees.

Provide an example of both pre-pruning and of post-pruning.

8. Give a small classification problem and apply the 1R algorithm to it.

9. Give the pseudo-code of a simple covering algorithm for classification rules.

10.How can the “Naïve Bayes” algorithm deal with missing values?

Veel succes!!!!

Good Luck!!!!