Support Vector Machine

Support Vector Machine is a powerful (but very slow) 'black-box' algorithm for classification.

Through the use of kernel functions, it can learn complex non-linear decision boundaries (i.e., when it is not possible to compute the {{ isCausalPrediction() | targetRoleName}} as a linear combination of input features).

SVM is effective with a large number of features. However, this algorithm is generally much slower than others and is generally not practical with more than a few thousand records.

Kernel

Try with a linear kernel Try with a RBF kernel Try with a polynomial kernel Try with a sigmoid kernel

Gamma

Try with 1/nb_features (auto) Try with 1/(nb_features * variance) (scale) Try with custom gamma values

Kernel coefficient for RBF, polynomial and sigmoid kernels.
Gamma defines the 'influence' of each training example in the features space.
A low value of gamma means that each example has 'far-reaching influence', while a high value means that each example only has close-range influence.

Coef0

Independent term for 'polynomial' or 'sigmoid' kernel function.

Tolerance

Tolerance for stopping criterion, i.e. variation on loss below which the algorithm stops training. Should be a positive number.

Maximum number of iterations

-1 = unlimited