Xia Jiang

EAAI Journal 2025 Journal Article

A Multi-Scale Sparse Channel Transformer Network for image reconstruction of astronomical bright source contamination

• Yajuan Zhang
• Congcong Shen
• Xia Jiang
• Bo Qiu
• Ali Luo
• Fuji Ren
• Yuanlu Chen

ICLR Conference 2025 Conference Paper

DRoC: Elevating Large Language Models for Complex Vehicle Routing via Decomposed Retrieval of Constraints

• Xia Jiang
• Yaoxin Wu
• Chenhao Zhang
• Yingqian Zhang 0001

This paper proposes Decomposed Retrieval of Constraints (DRoC), a novel framework aimed at enhancing large language models (LLMs) in exploiting solvers to tackle vehicle routing problems (VRPs) with intricate constraints. While LLMs have shown promise in solving simple VRPs, their potential in addressing complex VRP variants is still suppressed, due to the limited embedded internal knowledge that is required to accurately reflect diverse VRP constraints. Our approach mitigates the issue by integrating external knowledge via a novel retrieval-augmented generation (RAG) approach. More specifically, the DRoC decomposes VRP constraints, externally retrieves information relevant to each constraint, and synergistically combines internal and external knowledge to benefit the program generation for solving VRPs. The DRoC also allows LLMs to dynamically select between RAG and self-debugging mechanisms, thereby optimizing program generation without the need for additional training. Experiments across 48 VRP variants exhibit the superiority of DRoC, with significant improvements in the accuracy rate and runtime error rate delivered by the generated programs. The DRoC framework has the potential to elevate LLM performance in complex optimization tasks, fostering the applicability of LLMs in industries such as transportation and logistics.

NeurIPS Conference 2025 Conference Paper

Large Language Models as End-to-end Combinatorial Optimization Solvers

• Xia Jiang
• Yaoxin Wu
• Minshuo Li
• Zhiguang Cao
• Yingqian Zhang

Combinatorial optimization (CO) problems, central to decision-making scenarios like logistics and manufacturing, are traditionally solved using problem-specific algorithms requiring significant domain expertise. While large language models (LLMs) have shown promise in automating CO problem solving, existing approaches rely on intermediate steps such as code generation or solver invocation, limiting their generality and accessibility. This paper introduces a novel framework that empowers LLMs to serve as end-to-end CO solvers by directly mapping natural language problem descriptions to solutions. We propose a two-stage training strategy: supervised fine-tuning (SFT) imparts LLMs with solution construction patterns from domain-specific solvers, while a feasibility-and-optimality-aware reinforcement learning (FOARL) process explicitly mitigates constraint violations and refines solution quality. Evaluation across seven NP-hard CO problems shows that our method achieves a high feasibility rate and reduces the average optimality gap to 1. 03–8. 20% by tuning a 7B-parameter LLM, surpassing both general-purpose LLMs (e. g. , GPT-4o), reasoning models (e. g. , DeepSeek-R1), and domain-specific heuristics. Our method establishes a unified language-based pipeline for CO without extensive code execution or manual architectural adjustments for different problems, offering a general and language-driven alternative to traditional solver design while maintaining relative feasibility guarantees.

AIIM Journal 2017 Journal Article

An algorithm for direct causal learning of influences on patient outcomes

• Chandramouli Rathnam
• Sanghoon Lee
• Xia Jiang

AAAI Conference 2006 Conference Paper

A Bayesian Network for Outbreak Detection and Prediction

• Xia Jiang

Health care officials are increasingly concerned with knowing early whether an outbreak of a particular disease is unfolding. We often have daily counts of some variable that are indicative of the number of individuals in a given community becoming sick each day with a particular disease. By monitoring these daily counts we can possibly detect an outbreak in an early stage. A number of classical time-series methods have been applied to outbreak detection based on monitoring daily counts of some variables. These classical methods only give us an alert as to whether there may be an outbreak. They do not predict properties of the outbreak such as its size, duration, and how far we are into the outbreak. Knowing the probable values of these variables can help guide us to a cost-effective decision that maximizes expected utility. Bayesian networks have become one of the most prominent architectures for reasoning under uncertainty in artificial intelligence. We present an intelligent system, implemented using a Bayesian network, which not only detects an outbreak, but predicts its size and duration, and estimates how far we are into the outbreak. We show results of investigating the performance of the system using simulated outbreaks based on real outbreak data. These results indicate that the system shows promise of being able to predict properties of an outbreak.