Summary of Training Green Ai Models Using Elite Samples, by Mohammed Alswaitti et al.
Training Green AI Models Using Elite Samples
by Mohammed Alswaitti, Roberto Verdecchia, Grégoire Danoy, Pascal Bouvry, Johnatan Pecero
First submitted to arxiv on: 19 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)
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 Medium Difficulty summary: This paper tackles the significant environmental implications of AI model training, emphasizing the need for energy-efficient and sustainable practices. The authors explore two approaches: data-centric methods for training energy-efficient models and instance selection methods that minimize training sets with negligible performance degradation. However, they note that the impact of data-centric training set selection on energy efficiency remains unexplored. To address this gap, the paper proposes an evolutionary-based sampling framework to identify elite training samples tailored to datasets and model pairs. The framework is evaluated using 8 AI classification models and 25 publicly available datasets, showing a 50% performance improvement and remarkable energy savings of 98% compared to traditional practices. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Low Difficulty summary: This paper talks about the environmental impact of creating artificial intelligence (AI) models. As more AI models are being created, it’s becoming a problem for our planet. The authors look at two ways to make AI model training better for the environment: using less data and selecting the best samples. However, they noticed that nobody has studied how these methods affect energy usage. To fix this, they came up with a new way to pick the best training samples based on evolution. They tested it with many different AI models and datasets, showing that it can make AI model training 50% better and use much less energy. |
Keywords
* Artificial intelligence * Classification
