I am trying to apply the machine learning model Support Vector Machine

library(caret)
library(e1071)
library(rpart)
library(mlbench)
##using the Glass data from the mlbench package
data(Glass, package = "mlbench")
##creating a training and a testing set
inTrain <- createDataPartition(y=Glass$Type, p=0.75, list=FALSE)
training <- Glass[inTrain,]
testing <- Glass[-inTrain,]

Here I prepare training scheme. Doing a repeated cross-validation with 10 folds and 3 repeats

control <- trainControl(method="repeatedcv", number=10, repeats=3)
# train the model using support vector machines radial basis function and using the above control
model <- train(Type~., data=training, method="svmRadial", trControl=control, tuneLength=5)
# summarize the model
print(model)

## Support Vector Machines with Radial Basis Function Kernel 
## 
## 162 samples
##   9 predictor
##   6 classes: '1', '2', '3', '5', '6', '7' 
## 
## No pre-processing
## Resampling: Cross-Validated (10 fold, repeated 3 times) 
## Summary of sample sizes: 147, 146, 147, 145, 147, 146, ... 
## Resampling results across tuning parameters:
## 
##   C     Accuracy   Kappa      Accuracy SD  Kappa SD 
##   0.25  0.6250136  0.4456794  0.08526855   0.1275448
##   0.50  0.6352315  0.4616255  0.07896820   0.1189824
##   1.00  0.6985321  0.5670292  0.09573832   0.1407493
##   2.00  0.7047522  0.5767124  0.09160754   0.1346595
##   4.00  0.7084259  0.5842125  0.07855055   0.1165467
## 
## Tuning parameter 'sigma' was held constant at a value of 0.3044607
## Accuracy was used to select the optimal model using  the largest value.
## The final values used for the model were sigma = 0.3044607 and C = 4.

I choose C=1 based on the cross-validation results above to train my svm model

svm.model <- svm(Type ~ ., data = training, cost = 1, gamma = 1)
##using the model to predict the testing data
svm.pred <- predict(svm.model, testing[,-10])
##compute the confusion matrix using the prediction and the true values'
confusionMatrix(svm.pred, testing[,10])

## Confusion Matrix and Statistics
## 
##           Reference
## Prediction  1  2  3  5  6  7
##          1 13  5  2  0  0  0
##          2  4 14  2  3  1  2
##          3  0  0  0  0  0  0
##          5  0  0  0  0  0  0
##          6  0  0  0  0  1  0
##          7  0  0  0  0  0  5
## 
## Overall Statistics
##                                           
##                Accuracy : 0.6346          
##                  95% CI : (0.4896, 0.7638)
##     No Information Rate : 0.3654          
##     P-Value [Acc > NIR] : 7.296e-05       
##                                           
##                   Kappa : 0.461           
##  Mcnemar's Test P-Value : NA              
## 
## Statistics by Class:
## 
##                      Class: 1 Class: 2 Class: 3 Class: 5 Class: 6 Class: 7
## Sensitivity            0.7647   0.7368  0.00000  0.00000  0.50000  0.71429
## Specificity            0.8000   0.6364  1.00000  1.00000  1.00000  1.00000
## Pos Pred Value         0.6500   0.5385      NaN      NaN  1.00000  1.00000
## Neg Pred Value         0.8750   0.8077  0.92308  0.94231  0.98039  0.95745
## Prevalence             0.3269   0.3654  0.07692  0.05769  0.03846  0.13462
## Detection Rate         0.2500   0.2692  0.00000  0.00000  0.01923  0.09615
## Detection Prevalence   0.3846   0.5000  0.00000  0.00000  0.01923  0.09615
## Balanced Accuracy      0.7824   0.6866  0.50000  0.50000  0.75000  0.85714