Summary of Many-objective Multi-solution Transport, by Ziyue Li et al.
Many-Objective Multi-Solution Transport
by Ziyue Li, Tian Li, Virginia Smith, Jeff Bilmes, Tianyi Zhou
First submitted to arxiv on: 6 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary Many-objective multi-solution Transport (MosT) is a framework that finds multiple diverse solutions in the Pareto front of many objectives. The approach optimizes weighted objectives for each solution using optimal transport between objectives and solutions. MosT ensures convergence to Pareto stationary solutions for complementary subsets of objectives, outperforming strong baselines on applications in federated learning, multi-task learning, and mixture-of-prompt learning for LLMs. This framework is critical for optimizing the performance of many objectives jointly with a few Pareto stationary solutions, which is essential in machine learning. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Many-objective multi-solution Transport (MosT) is a new way to solve problems that involve many goals or tasks. It finds different solutions that are good at achieving specific combinations of these goals. MosT does this by using an idea called “optimal transport” to balance the importance of each goal for each solution. This helps MosT find solutions that work well together and cover all the goals, which is important in many areas like learning from multiple sources or tasks. |
Keywords
* Artificial intelligence * Federated learning * Machine learning * Multi task * Prompt
