Summary of Active Learning Of Deep Neural Networks Via Gradient-free Cutting Planes, by Erica Zhang et al.
Active Learning of Deep Neural Networks via Gradient-Free Cutting Planes
by Erica Zhang, Fangzhao Zhang, Mert Pilanci
First submitted to arxiv on: 3 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Optimization and Control (math.OC)
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 paper introduces a novel gradient-free cutting-plane training method for ReLU networks of arbitrary depth, which is applied to develop an active learning scheme. This scheme achieves convergence guarantees with a geometric contraction rate of the feasible set, outperforming popular deep active learning baselines on both synthetic data and real-world sentimental classification tasks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper creates a new way to train ReLU networks using a cutting-plane method that doesn’t need gradients. It’s then used for an active learning approach that can guarantee it will converge. The results show this method is better than other popular deep learning approaches on both made-up data and real-world sentiment analysis tasks. |
Keywords
» Artificial intelligence » Active learning » Classification » Deep learning » Relu » Synthetic data
