Skip to the content.



G2MILP is the first deep generative framework for mixed-integer linear programming (MILP) instances. This framework can learn to generate novel and realistic MILP instances without prior expert-designed formulations, while preserving the structures and computational hardness of real-world datasets, simultaneously. Additionally, it is employed to generate challenging instances for benchmarking and research purposes. G2MILP offers a data-centric approach to enhancing MILP solver development, particularly in scenarios with limited data availability.

model architecture

Contact Us

Authors: Zijie Geng, Xijun Li, Jie Wang*.

Affiliation: MIRA Lab, University of Science and Technology of China / Noah’s Ark Lab, Huawei

Please contact:

“A Deep Instance Generative Framework for MILP Solvers Under Limited Data Availability”. Zijie Geng, Xijun Li, Jie Wang, Xiao Li, Yongdong Zhang, Feng Wu. NeurIPS 2023 (Spotlight). [paper]

“G2MILP: Learning to Generate Mixed-Integer Linear Programming Instances for MILP Solvers”. Jie Wang, Zijie Geng, Xijun Li, Jianye Hao, Yongdong Zhang, Feng Wu. [paper]

“Machine Learning Insides OptVerse AI Solver: Design Principles and Applications”. Xijun Li, Fangzhou Zhu, Hui-Ling Zhen, Weilin Luo, Meng Lu, Yimin Huang, Zhenan Fan, Zirui Zhou, Yufei Kuang, Zhihai Wang, Zijie Geng, Yang Li, Haoyang Liu, Zhiwu An, Muming Yang, Jianshu Li, Jie Wang, Junchi Yan, Defeng Sun, Tao Zhong, Yong Zhang, Jia Zeng, Mingxuan Yuan, Jianye Hao, Jun Yao, Kun Mao. [paper]