Summary of Combining Statistical Depth and Fermat Distance For Uncertainty Quantification, by Hai-vy Nguyen et al.
Combining Statistical Depth and Fermat Distance for Uncertainty Quantification
by Hai-Vy Nguyen, Fabrice Gamboa, Reda Chhaibi, Sixin Zhang, Serge Gratton, Thierry Giaccone
First submitted to arxiv on: 12 Apr 2024
Categories
- Main: Machine Learning (stat.ML)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Probability (math.PR); Applications (stat.AP)
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 a novel approach to measuring out-of-domain uncertainty in neural network predictions using “Lens Depth” (LD) combined with Fermat Distance. The method, which has no trainable parameters, captures the depth of a point with respect to a distribution in feature space without assuming the form of the distribution. The technique is applicable to any classification model and does not intervene in the training process, allowing it to provide excellent uncertainty estimation on toy datasets and competitive results on standard deep learning datasets compared to strong baseline methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper finds new ways to predict how well neural networks will do when faced with unknown situations. They use a special math tool called “Lens Depth” (LD) that helps figure out where something is in relation to other things, without assuming what those things look like. This method doesn’t need any training and can be used with any type of prediction model. It even works well on simple test cases and does just as well as or better than other popular methods. |
Keywords
* Artificial intelligence * Classification * Deep learning * Neural network
