Optimization

TMAC currently supports solving convex optimization problems with operator splitting methods. We provided a few examples, such as gradient descent for least squares, forward-backward splitting for regularized empirical risk minimization.

• Least squares
• Huber loss
• Regularized regression
• Intersection of convex sets
• Quadratically constrained quadratic program
• Total variation for image processing
• Portfolio optimization
• Support vector machine (SVM)