Tests on artificial data with 60% 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("data060.train", header=TRUE, sep="\t", quote="", row.names=1)
test  = read.table("data060.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
869	A	1.A1	2.A1	3.A1	4.A1	5.A6	6.C3	7.C4	8.A2	9.A1	⋯	11.A2	12.A1	13.B7	14.C3	15.A1	16.B2	17.A1	18.C6	19.A3	20.C3
19	A	1.A1	2.A1	3.B2	4.A1	5.A3	6.A7	7.A5	8.B1	9.A1	⋯	11.C5	12.B1	13.C1	14.A1	15.A1	16.A1	17.B4	18.A2	19.A4	20.C3
439	B	1.B1	2.B1	3.A1	4.B1	5.C1	6.B7	7.C1	8.C2	9.B1	⋯	11.A2	12.B2	13.B1	14.C1	15.B1	16.C1	17.B3	18.B1	19.B4	20.C2
870	A	1.C3	2.B3	3.C3	4.A1	5.C2	6.A5	7.C4	8.A2	9.A1	⋯	11.A1	12.A1	13.A1	14.B1	15.A1	16.A5	17.A1	18.A2	19.A5	20.C2
546	A	1.A1	2.C7	3.B2	4.A1	5.A3	6.C1	7.A2	8.C1	9.A1	⋯	11.A1	12.B2	13.A1	14.A1	15.A1	16.A5	17.B7	18.B1	19.C1	20.C2
567	B	1.A1	2.B2	3.B2	4.A1	5.B2	6.B3	7.B6	8.B1	9.B2	⋯	11.A1	12.A1	13.C1	14.C1	15.A1	16.C1	17.A1	18.B1	19.A3	20.C1
739	B	1.B4	2.C1	3.C3	4.B1	5.C1	6.B1	7.C4	8.C2	9.A1	⋯	11.B3	12.C3	13.B2	14.B1	15.A1	16.B2	17.B6	18.B1	19.B2	20.B1
437	A	1.A1	2.A1	3.A1	4.C2	5.A4	6.A4	7.A2	8.A1	9.A1	⋯	11.A1	12.C3	13.A1	14.A1	15.C3	16.C1	17.B4	18.A4	19.A4	20.B1
640	C	1.A1	2.C7	3.C1	4.C2	5.C1	6.C3	7.C3	8.C2	9.B1	⋯	11.A1	12.C2	13.B1	14.C3	15.C4	16.A1	17.A1	18.C2	19.C7	20.C1
440	B	1.B5	2.B2	3.B2	4.B1	5.A3	6.B2	7.A5	8.C2	9.C1	⋯	11.B4	12.B2	13.B1	14.B1	15.A1	16.B2	17.B7	18.B1	19.B4	20.C1

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)

97.1111111111111

85

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

94

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)

97.7777777777778

98

simpel.go

out <- system2(c("./simpel", "data060.train", "data060.test"), stdout=TRUE, stderr=TRUE)
cat(out, sep="\n")

Training score:	 98.0%
Testing score:	 98.0%