Cortex AI logo placeholderCortexLabAI · learning · engineering
Model Atlas · Margin methods

Support Vector Machine

Finds a maximum-margin decision boundary and can use kernels for nonlinear separation.

core mathematical viewmin ½||w||² + CΣξᵢ
Mental model

Understand it before memorizing it.

Choose the boundary that separates classes with the widest safe gap.

Best fit

Where this model earns its place

Medium datasets

Start with a simpler baseline then compare this model using the same evaluation split and operational constraints.

High-dimensional features

Start with a simpler baseline then compare this model using the same evaluation split and operational constraints.

Text classification

Start with a simpler baseline then compare this model using the same evaluation split and operational constraints.

Strengths and limits

Trade-offs matter more than popularity.

Strengths

✓ Strong margins

✓ Effective in high dimension

Limitations

△ Scaling can be expensive

△ Probability outputs are secondary

Evaluation

Metrics to watch

F1Interpret with the product objective and error cost.
Margin violationsInterpret with the product objective and error cost.
LatencyInterpret with the product objective and error cost.
Production checklist

Before it reaches real users

  1. 01

    Scale features

    Document the assumption and instrument the condition so regressions can be detected.

  2. 02

    Tune C/kernel

    Document the assumption and instrument the condition so regressions can be detected.

  3. 03

    Measure inference cost

    Document the assumption and instrument the condition so regressions can be detected.