Question

我有一个训练数据集

    Out  Revolver   Ratio     Num ...
0   1    0.766127   0.802982  0   ...
1   0    0.957151   0.121876  1 
2   0    0.658180   0.085113  0 
3   0    0.233810   0.036050  3 
4   1    0.907239   0.024926  5

结果变量Out是二进制的，只取值0或1. Num不是因素

然后我尝试使用nnet运行caret。我想最终尝试nnGrid，但我只是想确保它先行：

nnTrControl=trainControl(method = "cv", classProbs = TRUE, summaryFunction = twoClassSummary, 
    number = 2,verboseIter = TRUE, returnData = FALSE, returnResamp = "all") 
#nnGrid = expand.grid(.size=c(1,4,7),.decay=c(0,0.001,0.1))

Outf<-factor(training$Out)
model <- train(Outf~ Revolver+Ratio+Num, data=training, method='nnet',
           trControl = nnTrControl, metric="logLoss")#, tuneGrid=nnGrid)

我收到错误

Error in train.default(x, y, weights = w, ...) : 
At least one of the class levels is not a valid R variable name; This will cause errors when class probabilities are generated because the variables names will be converted to  X0, X1 . Please use factor levels that can be used as valid R variable names  (see ?make.names for help).

但是，我之前使用过caret并收到此错误，我使用make.names解决了这个问题。所以当我尝试以下代替时：

yCat<-make.names(training$Out, unique=FALSE, allow_=TRUE)
mnn <- model.matrix( ~Revolver + Ratio + Num, data = training)
model <- train(y=yCat, x=mnn, method='nnet',
           trControl = nnTrControl, metric="logLoss")#, tuneGrid=nnGrid)

然后我收到消息

The metric "logLoss" was not in the result set. ROC will be used instead.

但我不明白为什么不根据logLoss进行评估？

如果我然后用它来预测测试集

probs<-predict(model, newdata=testSet, type="prob")

我得到了

 Error in eval(expr, envir, enclos) : object '(Intercept)' not found

我该如何解决这个问题？

Caret nnet：logloss不适用于twoClassSummary

0 个答案: