# dm3part2 - University of Florida CISE department...

This preview shows pages 1–9. Sign up to view the full content.

University of Florida CISE department Gator Engineering Classification Part 2 Dr. Sanjay Ranka Professor Computer and Information Science and Engineering University of Florida, Gainesville

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
University of Florida CISE department Gator Engineering Data Mining Sanjay Ranka Spring 2011 Overview • Rule based Classifiers • Nearest-neighbor Classifiers
University of Florida CISE department Gator Engineering Data Mining Sanjay Ranka Spring 2011 Rule Based Classifiers • Classify instances by using a collection of “if … then …” rules • Rules are presented in Disjunctive Normal Form, R = (r 1 v r 2 v … r k ) R is called rule set r i ’s are called classification rules • Each classification rule is of form r i : (Condition i ) y •Condition is a conjunction of attribute tests y is the class label

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
University of Florida CISE department Gator Engineering Data Mining Sanjay Ranka Spring 2011 Rule Based Classifiers r i : (Condition i ) y – LHS of the rule is called rule antecedent or pre-condition – RHS is called the rule consequent • If the attributes of an instance satisfy the pre- condition of a rule, then the instance is assigned to the class designated by the rule consequent • Example – (Blood Type=Warm) (Lay Eggs=Yes) Birds – (Taxable Income < 50K) (Refund=Yes) Cheat=No
University of Florida CISE department Gator Engineering Data Mining Sanjay Ranka Spring 2011 Classifying Instances with Rules • A rule r covers an instance x if the attributes of the instance satisfy the condition of the rule • Rule: r : (Age < 35) (Status = Married) Cheat=No • Instances: x 1 : (Age=29, Status=Married, Refund=No) x 2 : (Age=28, Status=Single, Refund=Yes) x 3 : (Age=38, Status=Divorced, Refund=No) • Only x 1 is covered by the rule r

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
University of Florida CISE department Gator Engineering Data Mining Sanjay Ranka Spring 2011 Rule Based Classifiers • Rules may not be mutually exclusive – More than one rule may cover the same instance • Strategies: – Strict enforcement of mutual exclusiveness • Avoid generating rules that have overlapping coverage with previously selected rules – Ordered rules • Rules are rank ordered according to their priority – Voting • Allow an instance to trigger multiple rules, and consider the consequent of each triggered rule as a vote for that particular class
University of Florida CISE department Gator Engineering Data Mining Sanjay Ranka Spring 2011 Rule Based Classifiers • Rules may not be exhaustive • Strategy: – A default rule r d : ( ) ˠ y d can be added – The default rule has an empty antecedent and is applicable when all other rules have failed y d is known as default class and is often assigned to the majority class

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
CISE department Gator Engineering Data Mining Sanjay Ranka Spring 2011 Example of Rule Based Classifier r 1 : (Refund=No) & (Marital Status=Single) & (Taxable Income>80K) ˠ Yes r 2
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 11/13/2011 for the course CIS 4930 taught by Professor Staff during the Spring '08 term at University of Florida.

### Page1 / 41

dm3part2 - University of Florida CISE department...

This preview shows document pages 1 - 9. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online