Tests on artificial data with 70% 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("data070.train", header=TRUE, sep="\t", quote="", row.names=1)
test  = read.table("data070.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
758	B	1.B2	2.C4	3.B1	4.C1	5.B2	6.B1	7.C1	8.A2	9.A1	⋯	11.B1	12.B2	13.A2	14.A1	15.B1	16.A1	17.A4	18.B3	19.B1	20.C1
987	A	1.B1	2.B1	3.C1	4.A1	5.C4	6.C1	7.C1	8.B6	9.B3	⋯	11.A1	12.A1	13.B1	14.A1	15.A4	16.B6	17.B7	18.C1	19.C1	20.A1
392	B	1.C1	2.C4	3.B1	4.A1	5.A1	6.B1	7.B2	8.C1	9.A1	⋯	11.B2	12.A1	13.A2	14.B3	15.C1	16.A1	17.A4	18.A2	19.B1	20.B1
820	C	1.A3	2.B1	3.C1	4.C1	5.A1	6.C1	7.C1	8.C1	9.C1	⋯	11.A1	12.C1	13.B4	14.C3	15.C1	16.A1	17.C2	18.C1	19.C1	20.A1
485	B	1.B2	2.A3	3.A1	4.A1	5.B2	6.B1	7.B1	8.B8	9.A1	⋯	11.B2	12.B2	13.A2	14.B4	15.B1	16.B5	17.B7	18.C1	19.A2	20.B2
251	C	1.C1	2.C4	3.A1	4.C1	5.B2	6.A1	7.A1	8.A2	9.B3	⋯	11.A1	12.C1	13.B5	14.C7	15.C1	16.B3	17.A3	18.C1	19.B1	20.A1
247	B	1.B1	2.A1	3.C1	4.B1	5.B2	6.C1	7.B1	8.B9	9.B3	⋯	11.C1	12.C1	13.B6	14.C7	15.B1	16.A1	17.B7	18.B3	19.C1	20.C1
882	C	1.B1	2.C2	3.B1	4.B1	5.C6	6.A1	7.C1	8.B9	9.C1	⋯	11.A1	12.C2	13.A2	14.C2	15.B1	16.C3	17.C2	18.B1	19.C1	20.C1
549	B	1.B2	2.A3	3.B1	4.B1	5.B2	6.B1	7.A1	8.B6	9.C1	⋯	11.B1	12.B1	13.B6	14.C4	15.B1	16.B6	17.B7	18.B3	19.B1	20.B2
584	B	1.B1	2.A1	3.B4	4.A4	5.C7	6.B1	7.A1	8.B9	9.B3	⋯	11.A1	12.B4	13.C2	14.C4	15.C1	16.B6	17.B7	18.C1	19.A2	20.B1

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)

93.7777777777778

83

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

88

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)

93

92

simpel.go

out <- system2(c("./simpel", "data070.train", "data070.test"), stdout=TRUE, stderr=TRUE)
cat(out, sep="\n")

Training score:	 92.3%
Testing score:	 91.0%