I have tried using the split function to split the dataset and iterating into them and predicting confusion matrix for each iteration, and taking mean of all confusion matri