Small Shifts, Large Gains: Unlocking Traditional TSP Heuristic Guided-Sampling via Unsupervised Neural Instance Modification

The Traveling Salesman Problem (TSP) is one of the most representative NP-hard problems in route planning and a long-standing benchmark in combinatorial optimization. Traditional heuristic tour constructors, such as Farthest or Nearest Insertion, are computationally efficient and highly practical, but their deterministic behavior limits exploration and often leads to local optima. In contrast, neural-based heuristic tour constructors alleviate this issue through guided-sampling and typically achieve superior solution quality, but at the cost of extensive training and reliance on ground-truth supervision, hindering their practical use. To bridge this gap, we propose TSP-MDF, a novel instance modification framework that equips traditional deterministic heuristic tour constructors with guided-sampling capability. Specifically, TSP-MDF introduces a neural-based instance modifier that strategically shifts node coordinates to sample multiple modified instances, on which the base traditional heuristic tour constructor constructs tours that are mapped back to the original instance, allowing traditional tour constructors to explore higher-quality tours and escape local optima. At the same time, benefiting from our instance modification formulation, the neural-based instance modifier can be trained efficiently without any ground-truth supervision, ensuring the framework maintains practicality. Extensive experiments on large-scale TSP benchmarks and real-world benchmarks demonstrate that TSP-MDF significantly improves the performance of traditional heuristics tour constructors, achieving solution quality comparable to neural-based heuristic tour constructors, but with an extremely short training time.

💡 Research Summary

**
The paper tackles a long‑standing limitation of classic deterministic heuristics for the Euclidean Traveling Salesman Problem (TSP), namely their inability to explore multiple candidate tours because the insertion order is fixed. While recent neural‑based tour constructors can perform guided sampling and achieve higher quality solutions, they require large amounts of labeled data or costly reinforcement‑learning training, and the training time is prohibitive for large instances.

To bridge this gap, the authors propose TSP‑MDF (Tour Sampling via Modified Instances with a Neural‑based Instance Modifier), a three‑stage framework that augments any existing deterministic heuristic (e.g., Farthest Insertion, Nearest Insertion) with a lightweight, unsupervised neural network that modifies the coordinates of the original TSP instance.

Key components

1. Instance Modifier (pre‑processing) – a simple MLP or Graph Neural Network that learns a small displacement vector for each node. The network is trained without ground‑truth tours using a loss that directly measures the improvement of the heuristic’s tour length on the original instance after mapping back from the modified instance:
     L(θ) = E_{s′∈S}