Summary of Uncertainty Quantification For Gradient-based Explanations in Neural Networks, by Mihir Mulye and Matias Valdenegro-toro
Uncertainty Quantification for Gradient-based Explanations in Neural Networks
by Mihir Mulye, Matias Valdenegro-Toro
First submitted to arxiv on: 25 Mar 2024
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 The proposed pipeline combines uncertainty estimation methods and explanation methods to quantify the explanation uncertainty of neural networks. By generating explanation distributions for the CIFAR-10, FER+, and California Housing datasets using Guided Backpropagation, this approach can provide confidence in the explanations by evaluating their coefficient of variation. The quality of these explanations is also assessed through modified pixel insertion/deletion metrics. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research paper proposes a way to understand why AI models make certain predictions. It’s crucial for making sure these models work correctly and providing useful insights into how they think. To do this, the authors combine two important techniques: estimating uncertainty and generating explanations. They test their approach on three datasets (CIFAR-10, FER+, and California Housing) to see if it can provide reliable information about why AI models are making certain predictions. |
Keywords
* Artificial intelligence * Backpropagation
