A suite of stochastic optimization methods written in julia for solving the empirical risk minimization problem. .
Is package in Julia for calculating approximate pseudoinverse of a matrix using randomized iterative methods.
Stochastic block BFGS method for solving the empirical risk minimization problems with a logistic loss and L2 regularizer
Inverse Random is a suite of randomized methods for inverting positive definite matrices implemented in MATLAB
Random Linear Lab is a lab for testing and comparing randomized methods for solving linear systems all implemented in MATLAB
quasi-Newton action constrained is a MATLAB implementation of the Newton-PCG for solving nonlinear unconstrained optimization problems that uses the block BFGS method as a preconditioner. The packages also contains implementations of the BFGS and LBFGS methods.
High Order Reverse Automatic Differentiation is a C++ implementation of reverse Automatic Differentiation routines that calculate the Hessian matrix, its sparsity pattern and the directional derivative of the Hessian matrix. It is an additional driver of the ADOL-C package. For a quick short-cut installation of both ADOL-C and HighOrderReverse, try ADOL-C_2.5.0_HighOrderReverse-1.0.tar.gz
Halley-Chebyshev with Automatic Derivatives is a C++ implementation of the Halley-Chebyshev methods with automatic derivatives by interfacing with the ADOL-C and HighOrderAD package. The implementation is completely tensor free as it calculates the directional derivative of the Hessian matrix for the necessary third-order information. This software comes with a Challenge ⇨ HalleyChebyChallenge.pdf ⇦