Midterm Exam Review


The exam will be conceptual, and covers all the lecture materials so far including frequent itemset mining (Apriori and FP Growth, closed pattern and max-pattern mining), classification (decision tree, Naïve Bayes, ensemble methods, model evaluation, SVM, regression, neural networks, kNN classification), and clustering (k-means, evaluation).    


The exam will be open-book and open-notes.  You will have the full class period (75 minutes).  The exam will be governed by the Emory Honor Code.


The questions will be short answer questions similar to the exercise questions on the textbook and the exercises we do in class. Below are some samples. 


Chap 6. 6.6, 6.9, 6.10

Chap 8. 8.7

Chap 9. 9.4, 9.5

Chap 10. 10.2