[JMLR (CCF-A)] PyPop7: A Pure-PYthon LibrarY for POPulation-based Black-Box Optimization (BBO), especially *Large-Scale* algorithm variants (from evolutionary computation, swarm intelligence, statistics, operations research, machine learning, mathematical optimization, meta-heuristics, auto-control etc.). [https://jmlr.org/papers/v25/23-0386.html]

A GPU-Accelerated First Order Solver for Convex Quadratic Programming

Rapid Optimization Library

The SAG (Stochastic Average Gradient) + SAGA (Accelerated) solver is an optimization algorithm used primarily in machine learning, specifically for logistic regression and linear support vector machines (SVMs) within libraries like scikit-learn. It is designed to be highly efficient for large datasets with many samples and features. Solver

LSQR is an iterative method for solving large, sparse, linear systems of equations and linear least-squares problems, including under- or over-determined and rank-deficient systems. It uses the Lanczos bidiagonalization process to provide a robust alternative to conjugate gradients, offering better numerical stability. Solver

The Conjugate Gradient (CG) method is an efficient iterative algorithm for solving large, sparse systems of linear equations where the matrix is symmetric and positive-definite. It finds the minimum of a quadratic function by generating conjugate search directions, ensuring convergence in at most steps for an matrix.Solver

Linear regression with the LBFGSB (Limited-memory Broyden-Fletcher-Goldfarb-Shanno BFGS) solver method is a numerical optimization method used to find the minimum of an objective function. It is a gradient descent algorithm that uses an approximation of the Hessian matrix to minimize the function.

A First-Order LP Solver Accelerated on Multiple GPUs

C51 Distributional DQN (v0.8) for bridge fleet maintenance optimization. Implements categorical return distributions (Bellemare et al., PMLR 2017) with 300x speedup via vectorized projection. Combines Noisy Networks, Dueling DQN, Double DQN, PER, and n-step learning. Validated on 200-bridge fleet: +3,173 reward in 83 min (25k episodes).

Deep Q-Network implementation for optimal bridge maintenance planning using Markov Decision Process formulation with vectorized parallel training. Based on Phase 3 (Vectorized DQN) from dql-maintenance-faster project.

Deep Q-Network (DQN) implementation for optimal maintenance planning of 100-bridge fleet infrastructure using advanced reinforcement learning techniques and vectorized parallel training.

A deep reinforcement learning system for optimizing bridge maintenance decisions across municipal infrastructure fleets, implementing cross-subsidy budget sharing and cooperative multi-agent learning.

Problem for the CEC 2013 Special Session on Large-Scale Real-Parameter Optimization

Network-wide estimation of traffic flow and travel time with data-driven macroscopic models

Compressed Radiation Treatment Planning [NeurIPS'24, MP'2025, PMB'23]

Operations research projects will be added in this repo

This repository provides practical implementations, examples, and insights into various optimization methods, making it easier to understand and apply these concepts.

Python implementation of a Genetic Algorithm to solve large scale binary knapsack problem

A ray-based library of Distributed POPulation-based OPtimization for Large-Scale Black-Box Optimization.

A flexible framework for SD-like algorithms in Julia.