Research map / Reinforcement learning / Visual and representation RL
SplitNet: A Reinforcement Learning Based Sequence Splitting Method for the MinMax Multiple Travelling Salesman Problem
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| transform single-TSP sequences into mTSP tours with SplitNet splitting and reforming | Random mTSP · Random mTSP, N=30, m=3 | maximum tour length: 2.86 -> 2.23 | Helped |
| transform single-TSP sequences into mTSP tours with SplitNet splitting and reforming | Random mTSP · Random mTSP, N=1000, m=3 | maximum tour length: 14.92 -> 8.06 | Helped |
| transform single-TSP sequences into mTSP tours with SplitNet splitting and reforming | Random mTSP · Random mTSP, N=1000, m=10 | maximum tour length: 2.82 -> 3.32 | Hurt |
| use multiple initial single-TSP sequences instead of one (ablation) | SplitNet · Random mTSP, N=50, m=5 | maximum tour length: 2.182 -> 2.082 | Helped |
| increase the number of sampled sequences from 30 to 40 (ablation) | SplitNet · Random mTSP, N=50, m=5 | maximum tour length: 2.082 -> 2.078 | Helped |
| use a decision region rather than unrestricted splitting (ablation) | SplitNet · Random mTSP, N=200 | maximum tour length: 2.62 -> 2.25 | Helped |
| restrict splitting to the decision-region center only (ablation) | SplitNet · Random mTSP, N=200 | maximum tour length: 2.25 -> 2.32 | Hurt |
| add LKH3 reforming of each split subsequence (ablation) | SplitNet · Random mTSP, N=50, m=5 | maximum tour length: 2.124 -> 2.082 | Helped |
Closest papers by meaning
- H-TSP: Hierarchically Solving the Large-Scale Traveling Salesman Problem (AAAI 2023 · similarity 0.88)
- Preference Optimization for Combinatorial Optimization Problems (ICML 2025 · similarity 0.87)
- DIMES: A Differentiable Meta Solver for Combinatorial Optimization Problems (NeurIPS 2022 · similarity 0.87)
- DualOpt: A Dual Divide-and-Optimize Algorithm for the Large-scale Traveling Salesman Problem (AAAI 2025 · similarity 0.87)
- DPN: Decoupling Partition and Navigation for Neural Solvers of Min-max Vehicle Routing Problems (ICML 2024 · similarity 0.86)
- MTL-KD: Multi-Task Learning Via Knowledge Distillation for Generalizable Neural Vehicle Routing Solver (NeurIPS 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.