Decision Trees - Unit 3 PDF

Decision trees  LEARNING DECISION TREES Age Competition Type Profit Old Yes Software Down Old No Software Down Old No Hardware Down Mid Yes Software Down Mid Yes Hardware Down Mi...

Decision trees  LEARNING DECISION TREES Age Competition Type Profit Old Yes Software Down Old No Software Down Old No Hardware Down Mid Yes Software Down Mid Yes Hardware Down Mid No Hardware Up Mid No Software Up New Yes Software Up New No Hardware Up New No Software Up Age Compt Ty Pro Old Y S/w Do Old N S/w Do Old N H/w Do Mid Y S/w Do Mid Y H/w Do Mid N H/w Up Mid N S/w Up New Y S/w Up New N H/w Up New N S/w Up Age Compt Ty Pro Old Y S/w Do Old N S/w Do Old N H/w Do Mid Y S/w Do Mid Y H/w Do Mid N H/w Up Mid N S/w Up New Y S/w Up New N H/w Up New N S/w Up Age Compt Ty Pro Old Y S/w Do Old N S/w Do Old N H/w Do Mid Y S/w Do Mid Y H/w Do Mid N H/w Up Mid N S/w Up New Y S/w Up New N H/w Up New N S/w Up Age Compt Ty Pro Old Y S/w Do Old N S/w Do Old N H/w Do Mid Y S/w Do Mid Y H/w Do Mid N H/w Up Mid N S/w Up New Y S/w Up New N H/w Up New N S/w Up Age Compt Ty Pro Old Y S/w Do Old N S/w Do Old N H/w Do Mid Y S/w Do Mid Y H/w Do Mid N H/w Up Mid N S/w Up New Y S/w Up New N H/w Up New N S/w Up Age Compt Ty Pro Old Y S/w Do Old N S/w Do Old N H/w Do Mid Y S/w Do Mid Y H/w Do Mid N H/w Up Mid N S/w Up New Y S/w Up New N H/w Up New N S/w Up Age Compt Ty Pro Old Y S/w Do Old N S/w Do Old N H/w Do Mid Y S/w Do Mid Y H/w Do Mid N H/w Up Mid N S/w Up New Y S/w Up New N H/w Up New N S/w Up  4/14 0.8113 = 0.0289 Standard Deviation Reduction: Based on the decrease in SD after a dataset is split on an attribute. Constructing a decision tree is all about finding attribute that returns the highest SD Reduction. =9.32 – 7.66 =1.66 =9.32 – 7.66 =1.66 Consider CV = 22% Mean (overcast) =(46+43+52+44) / 4 = 46.25 (46.3)