You are on page 1of 1

University of Engineering & Technology, Mardan

Department of Computer Software Engineering

Subject: Data Mining

Midterm Task

Q1. Of the following algorithms (6)

A. C4.5 decision tree


B. One-R
C. PRISM covering
D. 1-Nearest Neighbor
E. Naive Bayes
F. Linear Regression

1.1. Which one(s) are fast in training but slow in classification?


1.2. Which one(s) produce classification rules?
1.3. Which one(s) require discretization of continuous attributes before application?

Q2. Convert the following decision tree to a set of classification rules (if/than rules with
no “else”s). (3)

Q3. for the given data set (CPU), use at least three (any) classification algorithms
(preferably in Weka) . Present the comparison results in a table. Discuss and analyze your
results. (5+3+8)

You might also like