similar to: Assessing prediction performance of SVM using e1071 package

Dear R-list! I am using the e1071 package in R to solve a binary classification problem in a dataset of round 180 predictor variables (blood metabolites) of two groups of subjects (patients and healthy controls). I am confused regarding the correct way to assess the classification accuracy of the trained svm. (A) The svm command allows to specificy via the 'cross=k' parameter to specify a

1. This is _not_ a bug in R itself. Please don't use R's bug reporting system for contributed packages. 2. This is _not_ a bug in svm() in `e1071'. I believe you forgot to take sqrt. 3. You really should use the `tot.MSE' component rather than the mean of the `MSE' component, but this is only a very small difference. So, instead of spread[i] <- mean(mysvm$MSE), you

Hi: I am a newer in using R for data mining, and find the "e1071" pakage an excellent tool in doing data mining work! what frustrated me recently is that when I using the function "svm" and using the "cross=10" parameters, I got all the "accuracies" of the model greater than 1. Isn't that the accuracy should be smaller than 1? so I wander how, the

Full_Name: Noel O'Boyle Version: 2.1.0 OS: Debian GNU/Linux Sarge Submission from: (NULL) (131.111.8.96) (1) Description of error The 10-fold CV option for the svm function in e1071 appears to give incorrect results for the rmse. The example code in (3) uses the example regression data in the svm documentation. The rmse for internal prediction is 0.24. It is expected the 10-fold CV rmse

Hi R users, I'm trying to run a SVM - regression using e1071 package but the function svm() all the time apply a classification method rather than a regression. svm.m1 <- svm(st ~ ., data = train, cost = 1000, gamma = 1e-03) Parameters: SVM-Type: C-classification SVM-Kernel: radial cost: 1000 gamma: 0.001 Number of Support Vectors: 209

Hi. I'm trying to plot a pair of intertwined spirals and an svm that separates them. I'm having some trouble. Here's what I tried. > library(mlbench) > library(e1071) Loading required package: class > raw <- mlbench.spirals(200,2) > spiral <- data.frame(class=as.factor(raw$classes), x=raw$x[,1], y=raw$x[,2]) > m <- svm(class~., data=spiral) > plot(m,

svm.fit<-svm(as.factor(out) ~ ., data=all_h, method="C-classification", kernel="radial", cost=bestc, gamma=bestg, cross=10) # model fitting svm.pred<-predict(svm.fit, hh, decision.values = TRUE, probability = TRUE) # find the probability, but can not find. attr(svm.pred, "probabilities") > attr(svm.pred, "probabilities") 1 0 1 0 0 2 0

This could be a bug in the e1071 svm code, but maybe not -- I guess I'll send it here anyway. It's reproducible. > x <- seq (0.1,5,by=0.05) > y <- log(x) + rnorm (x, sd=0.2) > library(e1071) > m <- svm (x,y) Process R segmentation fault at Tue Nov 20 23:34:19 2001 > version _ platform i686-pc-linux-gnu arch i686 os

Hi, If I try to read the codes of functions in e1071 package, it gives me following error message. >library(e1071) > svm function (x, ...) UseMethod("svm") <environment: namespace:e1071> > predict.svm Error: Object "predict.svm" not found > Can someone help me on this how to read the codes of the functions in the e1071 package? Thanks. Raj

Hello all, this is the first time I'm using R and e1071 package and SVM multiclass (and I'm not a statistician)! I'm very confused, then. The goal is: I have a sentence with sunny; it will be classified as "yes" sentence; I have a sentence with cloud, it will be classified as "maybe"; I have a sentence with rainy il will be classified as "no". The

Dear all, after having calibrated a svm model through the svm() command of the e1071 package, is there a way to i) represent the modeled relationships between the y and X variables (response variable vs. predictors)? ii) rank the influence of the predictors used in the model? Right now I am more interested in regression models, but I guess this would be useful for classification too. Thank

Hello, I am using SVM under e1071 package for nu-regression with 18 parameters. The variables are ordered factors, factors, date or numeric datatypes. I use the linear kernel. It gives the following error that I cannot solve. I tryed debug, browser and all that stuff, but no way. The error is: Error in get(ctr, mode = "function", envir = parent.frame())(levels(x), :

I expected, that I will get the same prediction, if I multiply the weights for all classes with a constant factor, but I got different results. Please look for the following code. > library(e1071) > data(Glass, package = "mlbench") > index <- 1:nrow(Glass) > testindex <- sample(index, trunc(length(index)/5)) > testset <- Glass[testindex, ] > trainset <-

Hello all, I am using the e1071 SVM with the tune options for classification, which work pretty well, given the examples of using tune.svm function for classification. But I have not found any example to tune the SVM novelty detection (one-classification) parameters (gamma, cost, nu), for example this are some of the options I have tried with no success: obj<-tune(svm, x,y, type

Hi, I ran two svm models in R e1071 package: the first without cross-validation and the second with 10-fold cross-validation. I used the following syntax: #Model 1: Without cross-validation: > svm.model <- svm(Response ~ ., data=data.df, type="C-classification", > kernel="linear", cost=1) > predict <- fitted(svm.model) > cm <- table(predict,

Hi, I ran two svm models in R e1071 package: the first without cross-validation and the second with 10-fold cross-validation. I used the following syntax: #Model 1: Without cross-validation: > svm.model <- svm(Response ~ ., data=data.df, type="C-classification", > kernel="linear", cost=1) > predict <- fitted(svm.model) > cm <- table(predict,

I am accustomed to the LIBSVM package, which provides cross-validation on training with the -v option % svm-train -v 5 ... This does 5 fold cross validation while building the model and avoids over-fitting. But I don't see how to accomplish that in the e1071 package. (I learned that svm(... cross=5 ...) only _tests_ using cross-validation -- it doesn't affect the training.) Can

Dear all, I got a code for e1071 package in R for SVM regression. I have used *m$coefs* for extracting the coefficients but I am getting only 72 . How can I extract coefficients of the predictors set? Does it mean that I will get only 72 as *Number of Support Vectors: 72. * ** Thanks in advance Code: -------------- library(e1071) > # create data > x <- seq(0.1, 5,

Hi all, I'm using SVM to classify data (2 classes) and I get strange results : > model = svm(x, y, probability = TRUE) > pred = predict(model, x, decision.values = TRUE, probability = FALSE) > table(pred,y) y pred ctl nuc ctl 82 3 nuc 5 84 > pred 1 2 3 4 5 6 7 8 .... nuc nuc nuc nuc nuc nuc nuc ctl .... And now, with probabities computation :

Hello, I'm using the e1071 package for training an SVM. It seems to be working well. This question has two parts: 1) Once I've trained an SVM model, I want to USE it within R at a later date to predict various new data. I see the write.svm command, but don't know how to LOAD the model back in so that I can use it tomorrow. How can I do this? 2) I would like to add the