Summary of Emoe: Expansive Matching Of Experts For Robust Uncertainty Based Rejection, by Yunni Qu (1) et al.
EMOE: Expansive Matching of Experts for Robust Uncertainty Based Rejection
by Yunni Qu, James Wellnitz, Alexander Tropsha, Junier Oliva
First submitted to arxiv on: 3 Jun 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 A novel method, Expansive Matching of Experts (EMOE), leverages support-expanding and extrapolatory pseudo-labeling to enhance prediction and uncertainty-based rejection of out-of-distribution points. EMOE employs an expansive data augmentation technique that generates such instances in a latent space, along with an empirical trial approach to filter out augmented expansive points for pseudo-labeling. This method utilizes a diverse set of multiple base experts as pseudo-labelers on the augmented data, feeding into a shared MLP with multiple heads (one per expert). Compared to state-of-the-art methods, EMOE achieves superior performance on tabular data. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary EMOE is a new way to improve predictions and reject points that are outside what we’re expecting. It uses a combination of techniques to generate more examples that are like the ones we’ve seen before, but also unlike them in certain ways. Then it uses these extra examples to help figure out which real-life points don’t fit the pattern. This method is good at making predictions on tables of data. |
Keywords
» Artificial intelligence » Data augmentation » Latent space
