(Nonlinear) Optimization Library

This library aims to implement different mathematical optimization algorithms, such as regular and conjugate gradient descent. Mathematics is backed by Math.NET Numerics.

Gradient Descent Algorithms

• Resilient Error Gradient Descent

Conjugate Gradient Descent Algorithms

• Hager-Zhang ("CG_DESCENT")
• Polak-Ribière (supporting preconditioning)
• Fletcher-Reeves

Line Search Algorithms

• Secant
• Hager-Zhang with quadratic stepping

Cost Functions

• Residual Sum of Squares