Summary of Tarot: Targeted Data Selection Via Optimal Transport, by Lan Feng et al.
TAROT: Targeted Data Selection via Optimal Transport
by Lan Feng, Fan Nie, Yuejiang Liu, Alexandre Alahi
First submitted to arxiv on: 30 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
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 TAROT framework, grounded in optimal transport theory, aims to improve targeted data selection for deep learning tasks. Previous methods rely on influence-based greedy heuristics, which struggle with complex multimodal distributions and high-dimensional feature spaces. TAROT addresses these limitations by incorporating whitened feature distance to mitigate dominant feature bias and minimize the optimal transport distance between selected data and target domains. The framework is evaluated across multiple tasks, including semantic segmentation, motion prediction, and instruction tuning, consistently outperforming state-of-the-art methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary TAROT is a new way to pick the most important data points for certain machine learning tasks. This helps make these tasks better by ignoring some features that are more important than others. TAROT works well on different types of tasks, like image recognition and motion prediction. It even beats other popular methods in many cases! |
Keywords
» Artificial intelligence » Deep learning » Instruction tuning » Machine learning » Semantic segmentation
