SVM (Pegasos-Linear)ΒΆ

The documentation is generated based on the sources available at dotnet/machinelearning and released under MIT License.

Type: binaryclassifiertrainer Aliases: LinearSVM, svm Namespace: Microsoft.ML.Trainers.Online Assembly: Microsoft.ML.StandardLearners.dll Microsoft Documentation: SVM (Pegasos-Linear)


The idea behind support vector machines, is to map the instances into a high dimensional space in which instances of the two classes are linearly separable, i.e., there exists a hyperplane such that all the positive examples are on one side of it, and all the negative examples are on the other. After this mapping, quadratic programming is used to find the separating hyperplane that maximizes the margin, i.e., the minimal distance between it and the instances.


Name Short name Default Description
batchSize batch 1 Batch size
initWtsDiameter initwts 0 Init weights diameter
initialWeights initweights   Initial Weights and bias, comma-separated
lambda   0.001 Regularizer constant
noBias   False No bias
numIterations iter 1 Number of iterations
performProjection project False Perform projection to unit-ball? Typically used with batch size > 1.
shuffle shuf True Whether to shuffle for each training iteration
streamingCacheSize cache 1000000 Size of cache when trained in Scope