10 papers · 1 filter
Rethinking Learnability in Offline Data-driven Optimization
Chao Qian, Chen-Guang Wang, Rong-Xi Tan +1
Black-Box Optimization (BBO) has broad applications, while traditional algorithms such as evolutionary algorithms and Bayesian optimization face efficiency challenges as real-world…
GRIMIP: A General Framework for Instance-Specific Configuration of MIP Solvers Using LLMs
Yidong Luo, Xuemin Chen, Chenguang Wang +3
Configuring the hyperparameters of Mixed-integer programming (MIP) solvers is a high-dimensional, instance-dependent optimization problem where suboptimal settings can degrade solv…
A Unified Framework for Data-Free One-Step Sampling via Wasserstein Gradient Flows
Chenguang Wang, Tianshu Yu
We develop a unified theoretical framework for data-free one-step sampling from unnormalized target distributions based on Wasserstein gradient flows. For a broad class of standard…
Order Matters in Retrosynthesis: Structure-aware Generation via Reaction-Center-Guided Discrete Flow Matching
Chenguang Wang, Zihan Zhou, Lei Bai +1
Template-free retrosynthesis methods treat the task as black-box sequence generation, limiting learning efficiency, while semi-template approaches rely on rigid reaction libraries…
Incomplete Data, Complete Dynamics: A Diffusion Approach
Zihan Zhou, Chenguang Wang, Hongyi Ye +2
Learning physical dynamics from data is a fundamental challenge in machine learning and scientific modeling. Real-world observational data are inherently incomplete and irregularly…
Importance Weighted Score Matching for Diffusion Samplers with Enhanced Mode Coverage
Chenguang Wang, Xiaoyu Zhang, Kaiyuan Cui +3
Training neural samplers directly from unnormalized densities without access to target distribution samples presents a significant challenge. A critical desideratum in these settin…
Are Your Generated Instances Truly Useful? GenBench-MILP: A Benchmark Suite for MILP Instance Generation
Yidong Luo, Chenguang Wang, Dong Li +1
The proliferation of machine learning-based methods for Mixed-Integer Linear Programming (MILP) instance generation has surged, driven by the need for diverse training datasets. Ho…
Beyond the Heatmap: A Rigorous Evaluation of Component Impact in MCTS-Based TSP Solvers
Xuanhao Pan, Chenguang Wang, Chaolong Ying +2
The ``Heatmap + Monte Carlo Tree Search (MCTS)'' paradigm has recently emerged as a prominent framework for solving the Travelling Salesman Problem (TSP). While considerable effort…
Towards Principled Task Grouping for Multi-Task Learning
Chenguang Wang, Xuanhao Pan, Tianshu Yu
Multi-task learning (MTL) aims to leverage shared information among tasks to improve learning efficiency and accuracy. However, MTL often struggles to effectively manage positive a…
Efficient Training of Multi-task Neural Solver for Combinatorial Optimization
Chenguang Wang, Zhang-Hua Fu, Pinyan Lu +1
Efficiently training a multi-task neural solver for various combinatorial optimization problems (COPs) has been less studied so far. Naive application of conventional multi-task le…