General Information Metrics for Improving AI Model Training Efficiency
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929657983533056 |
|---|---|
| author | Xu, Jianfeng Liu, Congcong Tan, Xiaoying Zhu, Xiaojie Wu, Anpeng Wan, Huan Kong, Weijun Li, Chun Xu, Hu Kuang, Kun Wu, Fei |
| author_facet | Xu, Jianfeng Liu, Congcong Tan, Xiaoying Zhu, Xiaojie Wu, Anpeng Wan, Huan Kong, Weijun Li, Chun Xu, Hu Kuang, Kun Wu, Fei |
| contents | To address the growing size of AI model training data and the lack of a universal data selection methodology-factors that significantly drive up training costs -- this paper presents the General Information Metrics Evaluation (GIME) method. GIME leverages general information metrics from Objective Information Theory (OIT), including volume, delay, scope, granularity, variety, duration, sampling rate, aggregation, coverage, distortion, and mismatch to optimize dataset selection for training purposes. Comprehensive experiments conducted across diverse domains, such as CTR Prediction, Civil Case Prediction, and Weather Forecasting, demonstrate that GIME effectively preserves model performance while substantially reducing both training time and costs. Additionally, applying GIME within the Judicial AI Program led to a remarkable 39.56% reduction in total model training expenses, underscoring its potential to support efficient and sustainable AI development. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_02004 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | General Information Metrics for Improving AI Model Training Efficiency Xu, Jianfeng Liu, Congcong Tan, Xiaoying Zhu, Xiaojie Wu, Anpeng Wan, Huan Kong, Weijun Li, Chun Xu, Hu Kuang, Kun Wu, Fei Machine Learning Artificial Intelligence Information Theory To address the growing size of AI model training data and the lack of a universal data selection methodology-factors that significantly drive up training costs -- this paper presents the General Information Metrics Evaluation (GIME) method. GIME leverages general information metrics from Objective Information Theory (OIT), including volume, delay, scope, granularity, variety, duration, sampling rate, aggregation, coverage, distortion, and mismatch to optimize dataset selection for training purposes. Comprehensive experiments conducted across diverse domains, such as CTR Prediction, Civil Case Prediction, and Weather Forecasting, demonstrate that GIME effectively preserves model performance while substantially reducing both training time and costs. Additionally, applying GIME within the Judicial AI Program led to a remarkable 39.56% reduction in total model training expenses, underscoring its potential to support efficient and sustainable AI development. |
| title | General Information Metrics for Improving AI Model Training Efficiency |
| topic | Machine Learning Artificial Intelligence Information Theory |
| url | https://arxiv.org/abs/2501.02004 |