Summary of Predict-then-optimize by Proxy: Learning Joint Models Of Prediction and Optimization, By James Kotary et al.
Predict-Then-Optimize by Proxy: Learning Joint Models of Prediction and Optimization
by James Kotary, Vincenzo Di Vito, Jacob Christopher, Pascal Van Hentenryck, Ferdinando Fioretto
First submitted to arxiv on: 22 Nov 2023
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 This paper proposes an alternative method for optimizing decision processes with unknown parameters by learning optimal solutions directly from observable features. The approach, based on the Learning-to-Optimize paradigm, is generic and can employ various existing techniques. By predictive models, optimal solutions are learned without requiring handcrafted rules for backpropagation through optimization steps. Experimental evaluations show that several methods provide efficient, accurate, and flexible solutions to challenging problems in the Predict-Then-Optimize framework. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper introduces a new way to optimize decisions when you don’t know all the details. Right now, people use machine learning models to predict what’s missing, then solve an optimization problem to make the best decision. But this can be slow and requires special rules for each problem. This paper suggests a simpler approach: learn the best solution directly from the available information. This works with many existing techniques and can handle tricky problems. The results show that this method is good at finding efficient, accurate, and flexible solutions. |
Keywords
* Artificial intelligence * Backpropagation * Machine learning * Optimization
