org.apache.spark.mllib.optimization
Updater
Related Doc:
package optimization
Compute an updated value for weights given the gradient, stepSize, iteration number and regularization parameter.
Compute an updated value for weights given the gradient, stepSize, iteration number and regularization parameter. Also returns the regularization value regParam * R(w) computed using the *updated* weights.
- Column matrix of size dx1 where d is the number of features.
- Column matrix of size dx1 where d is the number of features.
- step size across iterations
- Iteration number
- Regularization parameter
A tuple of 2 elements. The first element is a column matrix containing updated weights, and the second element is the regularization value computed using updated weights.
:: DeveloperApi :: Class used to perform steps (weight update) using Gradient Descent methods.
For general minimization problems, or for regularized problems of the form min L(w) + regParam * R(w), the compute function performs the actual update step, when given some (e.g. stochastic) gradient direction for the loss L(w), and a desired step-size (learning rate).
The updater is responsible to also perform the update coming from the regularization term R(w) (if any regularization is used).