Wikipedia - Klasyfikacja

Dane bez obróbki

 GhostMiner 3.0
 Test 10x10CV stratified

Set1

Dane Vektory Cechy Klasy Majority Best
pkg1set1 681 833 4 32.1 kNN k=1 euclid 95.9 0.2
pkg2set1 454 689 6 35.2 kNN k=1 manhatan 81.1 0.5
pkg3set1 353 549 4 39.1 kNN k=auto euclid 90.0 0.3
pkg4set1 434 688 7 30.1 kNN k=1 euclid 72.2 0.6

Inne ciekawe wyniki:

pkg2set1 acc. [%] std.
kNN k=1 euclid 80.6 0.8
SVM gauss. auto 73.8 5.8
NaiveBayes 72.0 4.3
SVM lin. auto 71.7 6.3
pkg3set1 acc. [%] std.
kNN k=1 euclid 89.6 0.4
SVM gauss. auto 86.7 0.4

Set3

Dane Vektory Cechy Klasy Majority Best
pkg1set3 833 33865 4 28.6 kNN k=1 euclid 92.4 0.3
pkg2set3 689 43300 6 36.1 kNN k=1 euclid 67.9 0.4
pkg3set3 549 54330 4 38.4 kNN k=1 euclid 75.7 0.9
pkg4set3 688 82998 7 32.3 kNN k-1 euclid 54.6 0.6

Set4

Dane Vektory Cechy Klasy
pkg1set4 833 33865 833
pkg2set4 689 43300 29
pkg3set4 549 54330 45
pkg4set4 688 82998 51

Set2

Dane Vektory Cechy Klasy
pkg1set2 681 833 47
pkg2set2 454 689 27
pkg3set2 373 549 32
pkg4set2 481 688 50