Machine Learning for Data Mining

Trying additional solutions

Go back to the SVM model, Status, and click on the Expert tab. You can change the regularization parameter to something higher, as the model was consistent at 5; consider this an exercise. But, instead, we will change the Kernel type for our next run to the most complex Sigmoid type. It is not recommended to change the Bias value. But you can change Gamma for better results, and you can experiment on those values later. For now, we will keep them on default, and click on Run:

Click on the Analyze tab, and then click on Run. Here is our result:

The model is consistent, but as you can see, the accuracy percentage ...