Back to GallerySupport Vector Machine
Support Vector Machines construct a hyperplane that separates classes with the maximum margin (distance to the closest points, called support vectors). It uses the kernel trick to map data into higher dimensions for non-linear separation.
Key Parameters & Visual Influence:
• C (Regularization): Balances margin size and training errors. Small C maximizes margin width but allows some misclassifications. Large C penalizes errors heavily, forcing a complex, narrow boundary.
• Kernel & Gamma: RBF (Gaussian) uses radial functions, where Gamma sets the radius of influence of support vectors. High Gamma creates tight, localized 'pockets' around points.
Dataset Studio
CSV format: Headers in first row, last column = label/target. All values must be numeric.
Example: x1, x2, label
Hyperparameters
C (Regularization)Penalty strength for classification errors. Low C allows softer boundaries; high C forces hard boundaries.1.00
0.01100
KernelMathematical projection mapping data into higher dimensions: Linear, RBF, or Polynomial.
Learning RateStep size taken during weight optimization updates.0.0100
0.00010.5
Max EpochsMaximum training epochs to run optimization updates.200
502000
4 / 6 parameters active