Start hereWhy do only a few boundary points determine the classifier?
An SVM chooses a separating boundary with a wide margin. Points touching or violating the margin—the support vectors—determine the solution.
Building interactive view…
Technical lensFormalise what the visual is doing
Soft-margin SVM balances margin width and violations through C. Kernels replace dot products to create nonlinear boundaries in an implicit feature space.
Technical questionUse a tiny case to make the mechanism observable. Soft-margin SVM balances margin width and violations through C. Kernels replace dot products to create nonlinear boundaries in an implicit feature space. Verify one intermediate quantity, state change or mapping independently; then predict the consequence of this change: Increase C and watch the model trade a wider margin for fewer training errors.
Practitioner lensUse it responsibly
Scale inputs. Tune C and kernel parameters jointly. Probability estimates require additional calibration.
Transfer testTransfer this idea to a new example and justify each decision using this practitioner rule: Scale inputs. Tune C and kernel parameters jointly. Probability estimates require additional calibration. Then explain what should change if you deliberately test: Increase C and watch the model trade a wider margin for fewer training errors.
Worked explorationUse the visual as an experiment, not decoration
Draw two separable classes and candidate separating lines. Identify the line with the largest margin to the nearest points. Move a far-away point and notice the boundary barely changes; move a support vector and it changes.
Technical lens
Soft-margin SVM balances margin width and violations through C. Kernels replace dot products to create nonlinear boundaries in an implicit feature space.
Practitioner check
Scale inputs. Tune C and kernel parameters jointly. Probability estimates require additional calibration.
Prediction before interactionIncrease C and watch the model trade a wider margin for fewer training errors.
Exploration walkthroughTurn the interaction into an evidence trail
Draw two separable classes and candidate separating lines. Identify the line with the largest margin to the nearest points. Move a far-away point and notice the boundary barely changes; move a support vector and it changes. Before moving the control, state your prediction. After the visual changes, name the specific state, statistic, boundary or mapping that changed and explain why that change is consistent—or inconsistent—with your prediction.
- Record one observable quantity before the interaction and the same quantity afterwards.
- Change one factor at a time so the causal effect of the control is inspectable.
- Use an edge or failure case to discover where the concept stops behaving as the simple story suggests.
