香港理工大学：Artificial Neural Networks for Data Mining

Transformed data Input Feature Output Feature 151-1-11111 <1> < > 12345678 <-1.1.-1.1.-1.1.-1.1> <1> > =1> <-1.1.1.-1.-1.1.-1.1> =1> > <1> <-1.1.1.-1.-1.-1.1.1> 1> 1.-1.-1.-1.1.-1.1.1> 1>

16 Transformed Data Input Feature Output Feature 1 <1, -1, -1, -1, 1, 1, -1, 1> <1> 2 <-1, 1, -1, -1, 1, -1, 1, 1> <-1> 3 <-1, 1, -1, 1, -1, 1, -1, 1> <1> 4 <1, -1, 1, -1, -1, 1, -1, 1> <-1> 5 <-1, 1, 1, -1, -1, 1, -1, 1> <-1> 6 <-1, 1, -1, -1, 1, 1, -1, 1> <1> 7 <-1, 1, 1, -1, -1, -1, 1, 1> <-1> 8 <1, -1, -1, -1, 1, -1, 1, 1> <-1>

Record 1 Input Feature:<1,-1,-1,-1,1,1,-1,1> Output Feature: <1> Original Weight: <0,0,0,0,0,0,0, 0> Weight Change:<1,-1,-1,-1,1,1,-1,1> ° New Weight:<1,-1,-1,-1,1,1,-1,1>

17 Record 1 • Input Feature: <1, -1, -1, -1, 1, 1, -1, 1> • Output Feature: <1> • Original Weight: <0, 0, 0, 0, 0, 0, 0, 0> • Weight Change: <1, -1, -1, -1, 1, 1, -1, 1> • New Weight: <1, -1, -1, -1, 1, 1, -1, 1>

Record 2 Input Feature:<-1,1,-1,-1,1,-1,1,1> Output Feature: <-1> Original Weight:<1,-1,-1,-1,1,1,-1,1> Weight Change:<1,-1,1,1,-1,1,-1,-1> ° New Weight:<2,-2,0,0,0,2,-2,0>

18 Record 2 • Input Feature: <-1, 1, -1, -1, 1, -1, 1, 1> • Output Feature: <-1> • Original Weight: <1, -1, -1, -1, 1, 1, -1, 1> • Weight Change: <1, -1, 1, 1, -1, 1, -1, -1> • New Weight: <2, -2, 0, 0, 0, 2, -2, 0>

Record 3 Input Feature:<-1,1,-1,1,-1,1,-1,1> Output Feature: <1> Original Weight:<2,-2,0,0,0,2,-2,0> Weight Change:<-1,1,-1,1,-1,1,-1,1> ° New Weight:<1,-1,-1,1,-1,3,-3,1>

19 Record 3 • Input Feature: <-1, 1, -1, 1, -1, 1, -1, 1> • Output Feature: <1> • Original Weight: <2, -2, 0, 0, 0, 2, -2, 0> • Weight Change: <-1, 1, -1, 1, -1, 1, -1, 1> • New Weight: <1, -1, -1, 1, -1, 3, -3, 1>

Record 4 Input Feature:<1,-1,1,-1,-1,1,-1,1> Output Feature: <-1> Original Weight:<1,-1,-1,1,-1,3,-3,1> Weight Change:<-1,1,-1,1,1,-1,1,-1> ° New Weight:<0,0,-2,2,0,2,-2,0>

20 Record 4 • Input Feature: <1, -1, 1, -1, -1, 1, -1, 1> • Output Feature: <-1> • Original Weight: <1, -1, -1, 1, -1, 3, -3, 1> • Weight Change: <-1, 1, -1, 1, 1, -1, 1, -1> • New Weight: <0, 0, -2, 2, 0, 2, -2, 0>