Summary of Scaleviz: Scaling Visualization Recommendation Models on Large Data, by Ghazi Shazan Ahmad et al.
ScaleViz: Scaling Visualization Recommendation Models on Large Data
by Ghazi Shazan Ahmad, Shubham Agarwal, Subrata Mitra, Ryan Rossi, Manav Doshi, Vibhor Porwal, Syam Manoj Kumar Paila
First submitted to arxiv on: 27 Nov 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Human-Computer Interaction (cs.HC); 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 Automated visualization recommendations (vis-rec) are crucial for extracting insights from new datasets. Typically, vis-rec models calculate a large number of statistics and then use machine-learning models to score or classify multiple visualization choices. However, state-of-the-art models rely on expensive statistics, making them infeasible for large datasets due to prohibitively large computational time. This paper proposes a novel reinforcement-learning (RL) framework that identifies the best set of input statistics within a given time budget using a vis-rec model. We apply our technique to two state-of-the-art vis-rec models on three real-world datasets, achieving significant reductions in time-to-visualize with minimal error introduction. Our approach is about 10X times faster than baseline approaches introducing similar error. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper talks about how we can help people see important patterns and trends in new data by making good choices about what visuals to show. Normally, these kinds of models take a long time to work because they need to look at many numbers and statistics. But this paper proposes a new way that’s much faster without sacrificing too much accuracy. They tested their method on real-world datasets and found it works really well, saving 10 times the amount of time compared to other approaches. |
Keywords
» Artificial intelligence » Machine learning » Reinforcement learning
