Spark-optimizer

This is set of optimizer for spark in java which can be direct used for any of the Machine learning algorithm.These algorithms can be easily integrated with the existing machine learning library in Spark.The following algorithm are implemented.

Simple Gradient Descent:

This is a implementation of gradient descent with no regularization.

Conjugate Gradient Method

This is a implementation of the nonlinear conjugate gradient method to find the local minimum of function. For line search it uses StrongWolfe condition with bracketing and for interpolation in the zoom phase it uses Quadratic interpolation technique.