Templated unconstrained optimization library testing for simple optimization approaches:

gradient descent, newton's method, Quasi Newton, ...

Using self-contained Vector and Matrix classes for optimization as well as offering interfaces to Armadillo and Eigen linear algebra libraries.