Project 1.4: Mining with WEKA

Classification

The goal of this assignment is to educate us about decision tree classification, popular classification algorithms and the classification implementations provided by Weka.

Specific Tasks

Read DM chapter 4 and CI chapter 9.

Think about the dataset and attribute types and what, if any, transformations should be made to create a better decision tree classifier.

Be sure to use a subset of attributes in your classification training.

Run the ZeroR classifier on the dataset and record the results.

Run a decision tree classifier on the dataset with Weka and record the results.

Challenge: can you improve your classifier by using certain attributes or engaging in different preprocessing?

Due Tuesday, October 20 @ 11:59PM

On Wednesday October 21, you should be prepared to discuss your experience and results.

Be sure to commit and push the details of your clustering experiments to your git repo.

Grading Criteria

You must have committed and pushed the results of your experiments (see 'Specific Tasks' above) to your git repository. It's ok for your experiments to fail, but you must describe your process as outlined above. "I couldn't get Weka to work / open the data" is not a valid failure.