MachineLearning.Optimization.MinibatchGradientDescent

Minibatch Gradient Descent

Documentation

minibatchGradientDescent

Arguments

:: Model a

=> Int

seed

-> Int

batch size

-> R

learning rate, alpha

-> a

model to learn

-> R

epsilon

-> Int

max number of iters

-> Regularization

regularization parameter, lambda

-> Matrix

matrix of features, X

-> Vector

output vector, y

-> Vector

vector of initial weights, theta or w

-> (Vector, Matrix)

vector of weights and learning path

Minibatch Gradient Descent method implementation. See MachineLearning.Regression for usage details.