2D Kernel Classifier Demo — Single Panel + Loss Choice

Click to add labeled 2D points. After fitting, the same panel shows decision regions and the decision boundary for a kernel classifier of the form f(x)=∑ᵢ αᵢ k(xᵢ,x)+b. You can switch the kernel, the loss (logistic / hinge / squared / absolute on ±1 targets), and the optimizer settings.

class +1 class −1 decision boundary

Label to add

Domain

Kernel Loss Polynomial degree d: RBF gamma γ: Regularization λ: Learning rate η: Steps:

Suggested use
Load the concentric example and compare:
• linear + logistic
• poly degree 2 + logistic
• RBF + logistic

Then change the loss to hinge / squared / absolute to show how the fitted boundary changes. The γ and λ ranges are intentionally wide so you can show under/over-smoothing.

2D kernel classifier: add points and show the boundary in the same panel