2019 Lecture 17 Covering algorithms II 1 Covering Example continued 2 Age Spectacle prescription Astigmatism Tear production rate Recommended lenses Young Myope No Reduced None Young ID: 920862
Download Presentation The PPT/PDF document "Data Mining CSCI 307, Spring" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Data MiningCSCI 307, Spring 2019Lecture 17
Covering algorithms II
1
Slide2Covering Example continued
2
Age
Spectacle prescription
Astigmatism
Tear production rateRecommended lensesYoungMyopeNoReducedNoneYoungMyopeNoNormalSoftYoungMyopeYesReducedNoneYoungHypermetropeNoReducedNoneYoungHypermetropeNoNormalSoftYoungHypermetropeYesReducedNoneYoungHypermetropeYesNormalHardPre-presbyopicMyopeNoReducedNonePre-presbyopicMyopeNoNormalSoftPre-presbyopicMyopeYesReducedNonePre-presbyopicHypermetropeNoReducedNonePre-presbyopicHypermetropeNoNormalSoftPre-presbyopicHypermetropeYesReducedNonePre-presbyopicHypermetropeYesNormalNonePresbyopicMyopeNoReducedNonePresbyopicMyopeNoNormalNonePresbyopicMyopeYesReducedNonePresbyopicHypermetropeNoReducedNonePresbyopicHypermetropeNoNormalSoftPresbyopicHypermetropeYesReducedNonePresbyopicHypermetropeYesNormalNone
After first rule established, delete instances covered by the first rule and start again.
"Fresh" data set
Slide3Example part 2: Contact Lens DataRule we seek:Possible Tests:
3
Age = Young
Age = Pre-
presbyopic
Age = Presbyopic Spectacle prescription = Myope Spectacle prescription = HypermetropeAstigmatism = noAstigmatism = yesTear production rate = ReducedTear production rate = Normal
Slide4part 2: Modified Rule and Resulting DataRule with the best test is added:Instances covered by modified rule:
4
Age
Spectacle prescription
Astigmatism
Tear production rateRecommended lensesYoungMyopeNoReducedNoneYoungMyopeNoNormalSoftYoungMyopeYesReducedNoneYoungHypermetropeNoReducedNoneYoungHypermetropeNoNormalSoftYoungHypermetropeYesReducedNoneYoungHypermetropeYesNormalHard
Slide5Example part 2: RefineRule we seek:Possible Tests:
5
Spectacle prescription =
Myope
Spectacle prescription = HypermetropeAstigmatism = noAstigmatism = yesTear production rate = ReducedTear production rate = Normal
Slide6part 2: Modified Rule and Resulting DataRule with the best test is added:Instances covered by modified rule:
6
Age
Spectacle prescription
Astigmatism
Tear production rateRecommended lensesYoungMyopeYesReducedNoneYoungHypermetropeYesReducedNoneYoungHypermetropeYesNormalHard
Slide7Example part 2: Refine MoreRule we seek:Possible Tests:
7
Spectacle prescription =
Myope
Spectacle prescription = HypermetropeTear production rate = ReducedTear production rate = Normal
Slide8The Result
8
Final Rules:
Second rule for recommending "hard lenses":
(build from the instances not covered by the first rule.)
If astigmatism = yes and tear production rate = normal and spectacle prescription = myope then recommendation = hard
Slide9The Rest of the PRISM Rules
9
If
astigmatism = no
and
tear-prod-rate = normal and spectacle-prescription = hypermetrope then softIf astigmatism = no and tear-prod-rate = normal and age = young then softIf age = pre-presbyopic and astigmatism = no and tear-prod-rate = normal then softIf tear-prod-rate = reduced then noneIf age = presbyopic and tear-prod-rate = normal and spectacle-prescription = myope and astigmatism = no then noneIf spectacle-prescrip = hypermetrope and astigmatism = yes and age = pre-presbyopic then noneIf age = presbyopic and spectacle-prescription = hypermetrope and astigmatism = yes then none
Slide10Pseudo-Code for PRISM
10
Slide11Rules versus Decision ListsPRISM with outer loop removed generates a decision list for one classSubsequent rules are designed for rules that are not covered by previous rulesBut: order doesn’t matter because all rules predict the same classOuter loop considers all classes separatelyNo order dependence impliedProblems: overlapping rules, default rule required
11
Slide12Separate and ConquerMethods like PRISM (for dealing with one class) are separate-and-conquer algorithms:First, identify a useful ruleThen, separate out all the instances it coversFinally, “conquer” the remaining instances
12