Contents
What do you think a larger or smaller C value is good for the better classification performance SVM?
The larger C is the less the final training error will be. But if you increase C too much you risk losing the generalization properties of the classifier, because it will try to fit as best as possible all the training points (including the possible errors of your dataset).
What is the influence of C in SVMs with?
Tuning C correctly is a vital step in best practice in the use of SVMs, as structural risk minimisation (the key principle behind the basic approach) is party implemented via the tuning of C. The parameter C enforces an upper bound on the norm of the weights, which means that there is a nested set of hypothesis classes indexed by C.
How can we adjust the parameter c ( SVM )?
C is a trade-off between training error and the flatness of the solution. The larger C is the less the final training error will be. But if you increase C too much you risk losing the generalization properties of the classifier, because it will try to fit as best as possible all the training points (including the possible errors of your dataset).
Which is the better parameter to tune RBF-SVM?
A good way to estimate the better C to be used is to perform a grid search before the final training. Suppose you are using RBF-SVM, then you have two parameters to tune: C and gamma (the radius of RBF). C is a parameter controlling the importance of the training error respect to the margin:
Which is the best tool for visualizing SVM data?
Weka is a good tool for visualizing data and playing around with different settings for an SVM. It may help you get a better idea of how your data look and why changing the c value does not change the classification error. In general, having few training instances and many attributes make it easier to make a linear separation of the data.