Summary of From Inverse Optimization to Feasibility to Erm, by Saurabh Mishra et al.
From Inverse Optimization to Feasibility to ERM
by Saurabh Mishra, Anant Raj, Sharan Vaswani
First submitted to arxiv on: 27 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Optimization and Control (math.OC)
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 The paper studies contextual inverse optimization, which infers unknown parameters of an optimization problem from known solutions. It focuses on contextual inverse linear programming (CILP), addressing the challenges posed by non-differentiable LPs. The authors propose a reduction to a convex feasibility problem using alternating projections and provide theoretical convergence guarantees. They also reduce CILP to empirical risk minimization (ERM) on a smooth, convex loss, enabling the use of scalable first-order optimization methods. The paper quantifies generalization performance and experimentally validates its approach on synthetic and real-world problems, demonstrating improved performance compared to existing methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper is about using extra information to figure out how an optimization problem works when you only have the answers, not the questions. It looks at a special kind of linear programming problem that’s hard because it’s not smooth. The authors find a way to make it smoother and show that their method can be used with big problems. They test it on some fake data and real-world examples and show that it works better than other methods. |
Keywords
* Artificial intelligence * Generalization * Optimization
