Summary of Sourcerer: Sample-based Maximum Entropy Source Distribution Estimation, by Julius Vetter et al.
Sourcerer: Sample-based Maximum Entropy Source Distribution Estimation
by Julius Vetter, Guy Moss, Cornelius Schröder, Richard Gao, Jakob H. Macke
First submitted to arxiv on: 12 Feb 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 The proposed method targets the maximum entropy distribution to make a principled choice among many equally valid sources in scientific modeling applications. This approach prioritizes retaining as much uncertainty as possible and is suitable for simulators with intractable likelihoods, leveraging the Sliced-Wasserstein distance to measure the discrepancy between the dataset and simulations. The method outperforms recent source estimation methods in recovering source distributions with substantially higher entropy without sacrificing simulation fidelity. It is demonstrated on several tasks and used to infer source distributions for parameters of the Hodgkin-Huxley model from experimental datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Scientists often try to figure out what’s behind a set of data. This can be tricky because many different things could have caused that data. To solve this problem, a new method is proposed that tries to keep as much uncertainty as possible. It works by comparing the data with simulations and finding the one that best matches. This approach is good for situations where it’s hard to calculate how likely each simulation is. The method does better than other methods at keeping uncertainty while still matching the data well. It’s even used to figure out what’s behind experimental data about tiny brain cells. |
