search for: h_aspire

Displaying 1 result from an estimated 1 matches for "h_aspire".

Did you mean: aspire
2009 May 11
1
Problems to run SVM regression with e1071
...rnel: radial cost: 1000 gamma: 0.001 Number of Support Vectors: 209 I know that it is wrong because when I do prediction appear levels. I'm working with normalized data [0,1] (249 points) . I don't have idea what it is wrong. Somebody can help me? h_aspire dados=read.table("svmdata.txt",header=TRUE) index=1:nrow(d) test=d[210:249,] train=d[1:209,] require(e1071) tuneobj = tune.svm(st ~ ., data = train, gamma = 10^(-6:-3), cost = 10^(1:3summary(tuneobj) svm.m1 <- svm(st ~ ., data = train, cost = 1000, gamma = 1e-03) svm.pre...