SequentialMinimalOptimizationOptions
Options & Configuration · AiDotNet.Models.Options
Configuration for the Sequential Minimal Optimization solver used to train support-vector machines.
Inheritance: ModelOptions → SequentialMinimalOptimizationOptions
For Beginners
SMO trains an SVM by repeatedly fixing the two multipliers that are furthest from satisfying the optimality rules.
How It Works
Reference: Platt, “Sequential Minimal Optimization: A Fast Algorithm for Training Support Vector Machines” (1998).
Constructors
| Constructor | Summary |
|---|---|
SequentialMinimalOptimizationOptions() |
Initializes the options with documented defaults. |
SequentialMinimalOptimizationOptions(SequentialMinimalOptimizationOptions other) |
Initializes the options by copying another configuration. |
Properties
| Property | Summary |
|---|---|
bool RestrictPairsToSameLabel { get; set; } |
Gets or sets whether the working pair must be drawn from a single class. |
double StepEpsilon { get; set; } |
Gets or sets the smallest relative multiplier movement that counts as progress. |
double Tolerance { get; set; } |
Gets or sets how far a multiplier may violate the KKT conditions before it is considered for optimization. |
int MaxIterations { get; set; } |
Gets or sets the maximum number of multiplier pairs optimized. |