用默认参数得到的模型的最大准确度是80%。我现在用以下代码调参数：
grid_svmpoly=expand.grid(degree=seq(0,5,length.out=25),scale=seq(0,5,length.out=25),C=seq(0,5,length.out=25))
modelFit <- train(group~.,data=table_svm,method="svmPoly",
                  preProc = c("center","scale"),#metric="ROC",
                  tuneGrid=grid_svmpoly,
                  trControl = ctrl)

结果报错：Error in { :
  task 1 failed - "No Support Vectors found. You may want to change your parameters"

该如何解决呢？