This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Decision Trees CS 221 Section 5 October 23, 2009 Today we will work through an example problem of creating a decision tree. Time permitting, we will also work an example problem about AdaBoost. 1 Creating a decision tree Consider the criteria for accepting candidates to the Ph.D. program at the myth- ical University of St. Nordaf. Each candidate is evaluated according to four attributes: The grade point average (GPA), the quality of the undergraduate university attended, the publication record, and the strength of the recommen- dation letters. To simplify our example, let us discretize and limit the possible value of each attribute: Possible GPA scores are 4.0, 3.7, and 3.5; universities are categorized as top10, top20, and top30 (by top20 we mean places 11-20, and by top 30 we mean 21-30); publication record is a binary attribute - either the applicant has published previously or not; and recommendation letters are similarly binary, they are either good or normal. Finally, the candidates are classified into two classes: accepted, or P (for ‘positive’), and rejected, or N (for ‘negative’). Following is an example of one possible decision tree determining acceptance. 1 Applicant Pat doesn’t know this decision tree, but she does have the follow- ing data regarding twelve of last year’s applicants: Attributes No. GPA University Published Recommendation Class 1 4.0 top10 yes good P 2 4.0 top10 no good P 3 4.0 top20 no normal P 4 3.7 top10 yes good P 5 3.7 top20 no good P 6 3.7 top30 yes good P 7 3.7 top30 no good N 8 3.7 top10 no good N 9 3.5 top20 yes normal N 10 3.5 top10 no normal N 11 3.5 top30 yes normal N 12 3.5 top30 no good N 1. Verify that the tree given above correctly categorizes these examples. Answer: Yes the tree provided does classify the examples correctly. We can verify this by tracing a path from the root to a leaf for every example and confirming that the leaf value matches the example value....
View Full Document
- Artificial Intelligence, Decision tree learning, information gain