Tests on artificial data with 50% noise

library(adabag)
library(naivebayes)

Loading required package: rpart
Loading required package: caret
Loading required package: lattice
Loading required package: ggplot2
Loading required package: foreach
Loading required package: doParallel
Loading required package: iterators
Loading required package: parallel

# files available in: /net/aistaff/kleiweg/spraak/fa
train = read.table("data050.train", header=TRUE, sep="\t", quote="", row.names=1)
test  = read.table("data050.test",  header=TRUE, sep="\t", quote="", row.names=1)

train[1:10,]

	C.Class	C.W1	C.W2	C.W3	C.W4	C.W5	C.W6	C.W7	C.W8	C.W9	⋯	C.W11	C.W12	C.W13	C.W14	C.W15	C.W16	C.W17	C.W18	C.W19	C.W20
503	A	1.A2	2.C2	3.C1	4.A2	5.A1	6.A1	7.A1	8.A3	9.A4	⋯	11.B1	12.A1	13.A2	14.B5	15.A1	16.C2	17.A1	18.A2	19.A3	20.A1
554	C	1.A1	2.C1	3.C2	4.C1	5.A2	6.C1	7.C1	8.C1	9.B2	⋯	11.B1	12.C1	13.C8	14.C1	15.A2	16.C4	17.A1	18.C2	19.C8	20.C2
792	B	1.B4	2.B1	3.B1	4.B1	5.B1	6.B5	7.B1	8.C1	9.B2	⋯	11.B2	12.B2	13.B8	14.C1	15.B1	16.C3	17.A1	18.A3	19.B3	20.B1
743	B	1.B4	2.B1	3.B1	4.C1	5.B4	6.B5	7.B1	8.A3	9.B3	⋯	11.B1	12.B1	13.C7	14.A3	15.B1	16.A1	17.B7	18.B3	19.B3	20.B1
466	A	1.C3	2.A1	3.A1	4.B1	5.A1	6.B5	7.A1	8.A1	9.A5	⋯	11.B2	12.A1	13.B4	14.C1	15.A2	16.A1	17.A1	18.A3	19.B3	20.B1
542	C	1.C1	2.C2	3.A1	4.A1	5.C2	6.C1	7.B1	8.A1	9.C4	⋯	11.C1	12.C1	13.C3	14.B4	15.C1	16.C3	17.C1	18.A3	19.A5	20.C2
933	A	1.A1	2.B1	3.A1	4.A1	5.A2	6.A1	7.A1	8.C1	9.A1	⋯	11.B1	12.B4	13.A1	14.A2	15.A4	16.A1	17.A1	18.A3	19.C4	20.A1
440	B	1.C2	2.A1	3.B1	4.B1	5.A2	6.B5	7.B1	8.C1	9.B2	⋯	11.C1	12.C1	13.C4	14.C1	15.B1	16.B3	17.B5	18.C1	19.B2	20.B1
926	B	1.B3	2.B1	3.B1	4.B1	5.B2	6.B5	7.C1	8.C1	9.B3	⋯	11.B2	12.B3	13.B1	14.A4	15.B1	16.C4	17.B6	18.B4	19.C8	20.B1
403	C	1.B3	2.C1	3.C1	4.A1	5.B3	6.C1	7.C1	8.B6	9.C4	⋯	11.C2	12.C1	13.A1	14.C1	15.C1	16.B3	17.C1	18.C2	19.B2	20.A1

Bagging (AdaBag)

bag <- bagging(C.Class ~ ., data=train)
train.bagging <- predict(bag, newdata=train)
 test.bagging <- predict(bag, newdata=test)
100 * (1 - train.bagging$error)
100 * (1 -  test.bagging$error)

99.3333333333333

92

Boosting (AdaBoost)

boost <- boosting(C.Class ~ ., data=train)
train.boosting <- predict(boost, newdata=train)
 test.boosting <- predict(boost, newdata=test)
100 * (1 - train.boosting$error)
100 * (1 -  test.boosting$error)

100

96

Naive Bayes

score <- function(obs, exp) {
  return(100 * sum(obs == exp[,"C.Class"]) / length(obs))
}

nb <- naive_bayes(C.Class ~ ., data=train)
train.nb <- predict(nb, train)
 test.nb <- predict(nb, test)
score(train.nb, train)
score( test.nb, test)

99.8888888888889

98

simpel.go

out <- system2(c("./simpel", "data050.train", "data050.test"), stdout=TRUE, stderr=TRUE)
cat(out, sep="\n")

Training score:	 99.9%
Testing score:	 99.0%