Nonconvex embedded optimization: code generation for fast real-time optimization + ROS support

With Uno, finally take full control of your SQP/barrier solver for nonlinearly constrained optimization

Open-L2O: A Comprehensive and Reproducible Benchmark for Learning to Optimize Algorithms

Semi-random funky stuff, mainly for my PhD experiments and articles. Contains calculations and algorithm implementations for various applied mathematics and astrophysics articles I worked on.

Nonlinear programming application examples solved with Artelys Knitro

Simulation code for "Fast Converging Algorithm for Weighted Sum Rate Maximization in Multicell MISO Downlink" by Le-Nam Tran, Muhammad Fainan Hanif, Antti Tolli, Markku Juntti, IEEE Signal Processing Letters 19.12 (2012): 872-875

CaΣoS is a nonlinear sum-of-squares optimization suite based on the symbolic framework of CasADi.

Stochastic Gradient Descent (SGD) is an optimization algorithm that updates model parameters iteratively using small, random subsets (batches) of data, rather than the entire dataset. It significantly speeds up training for large datasets, though it introduces noise that causes, in some cases, heavy fluctuations.deep learning/neural networks.solver

Nonconvex Exterior Point Operator Splitting

Coordinate and Incremental Aggregated Optimization Algorithms

Build a Recurrent Neural Network solving Optimization Problems

No description provided.

Code for "A Regularized Newton Method for Nonconvex Optimization with Global and Local Complexity Guarantees"

Optimized and benchmarked parallel Genetic Algorithm with inequality constraints, and a scipy-like interface

HarvesterOpt is a Matlab code to maximize the energy recovered by a bistable energy harvester.

Solving Convex and Nonconvex Optimization Problems

Fix and Bound: An efficient approach for solving large-scale BoxQPs - Mathematical Programming Computation

Simulation code for "A Conic Quadratic Programming Approach to Physical Layer Multicasting for Large-Scale Antenna Arrays." by L.-N. Tran, M. F. Hanif, M. Juntti., IEEE Signal Processing Letters 21.1 (2014): 114-117

Quantum control algorithms for multiple examples based on the package Qutip

CEopt is a Matlab routine for non-convex optimization using the Cross-Entropy method and augmented Lagrangian formulation.

Perturbed Saddle-escape Descent (PSD): a first-order optimizer that escapes strict saddle points in nonconvex problems.

Code and results for quantum optimal control based on switching time optimization

CROSS-OPT is a Matlab package for optimizing truss structures with the Cross-Entropy method and augmented Lagrangian formulation.

Proof-of-concept implementation of a method by Song and Khan (2022) for computing convex relaxations for parametric ordinary differential equations.

R package for fast optimizing type I/II error, recall/precision and F1 score.

Investigating the Role of Weight Decay in Enhancing Nonconvex SGD, CVPR 2025

FraCTune is a Matlab package for tuning fractional-order controllers with the Cross-Entropy method and augmented Lagrangian formulation.

Numerical computation of invariant reachable sets for nonlinear systems using semi-algebraic programming

SpringpotTune is a Matlab package designed for fitting variable-order springpot models using the Cross-Entropy method.

Optimal Distributed Control of Multi-Agent Systems