Semi Stochastic Gradient Descenthttp://mloss.orgUpdates and additions to Semi Stochastic Gradient DescentenWed, 09 Jul 2014 04:28:47 -0000Semi Stochastic Gradient Descent 1.0<html><p>Efficient implementation of Semi-Stochastic Gradient Descent algorithm (S2GD) for training logistic regression (L2-regularized). </p> <p>The S2GD algorithm enjoys linear convergence (faster than SGD), and comes without the need of tuning parameters like stepsize (choice explained in paper). </p> <p>The code is in C++, called from MATLAB. The code is well commented, and should be easily adjusted for different applications. </p> <p>The package also contains implementation of SGD and SAG. </p> <p>First time usage: </p> <p>&gt;&gt; mexAll % compiles the .cpp files </p> <p>&gt;&gt; demo </p></html>jakub konecnyWed, 09 Jul 2014 04:28:47 -0000 gradient descentlogistic regression