计算机科学与探索 ›› 2023, Vol. 17 ›› Issue (1): 88-107.DOI: 10.3778/j.issn.1673-9418.2204019

• 前沿·综述 • 上一篇    下一篇

面向算法选择的元学习研究综述

李庚松,刘艺,秦伟,李红梅,郑奇斌,宋明武,任小广   

  1. 1. 国防科技创新研究院,北京 100071 
    2. 军事科学院,北京 100091
    3. 天津(滨海)人工智能创新中心,天津 300457
  • 出版日期:2023-01-01 发布日期:2023-01-01

Survey on Meta-Learning Research of Algorithm Selection

LI Gengsong, LIU Yi, QIN Wei, LI Hongmei, ZHENG Qibin, SONG Mingwu, REN Xiaoguang   

  1. 1. National Innovation Institute of Defense Technology, Beijing 100071, China
    2. Academy of Military Sciences, Beijing 100091, China
    3. Tianjin Artificial Intelligence Innovation Center, Tianjin 300457, China
  • Online:2023-01-01 Published:2023-01-01

摘要: 随着人工智能的快速发展,从可行的算法中选择满足应用需求的算法已经成为各领域亟待解决的关键问题,即算法选择问题。基于元学习的方法是解决算法选择问题的重要途径,被广泛应用于算法选择研究并取得了良好成果。方法通过构建问题特征到候选算法性能的映射模型来选择合适的算法,主要包括提取元特征、计算候选算法性能、构建元数据集以及训练元模型等步骤。首先,阐述基于元学习的算法选择概念和框架,回顾简述相关综述工作;其次,从元特征、元算法和元模型性能指标三方面总结研究进展,对其中典型的方法进行介绍并比较不同类型方法的优缺点和适用范围;然后,概述基于元学习的算法选择在不同学习任务中的应用情况;继而,使用140个分类数据集、9种候选分类算法和5种性能指标开展算法选择实验,对比不同算法选择方法的性能;最后,分析目前存在的挑战和问题,探讨未来的发展方向。

Abstract: With the rapid development of artificial intelligence, the selection of algorithms that meet application requirements from feasible algorithms has become a critical problem to be solved urgently in various fields, that is, the algorithm selection problem. The approach based on meta-learning is an important way to solve the algorithm selection problem, which is widely applied in algorithm selection research and achieves good results. The approach selects appropriate algorithms by constructing the mapping model from problem features to candidate algorithms performance, mainly including the steps of extracting meta-features, calculating candidate algorithms performance, constructing meta-dataset and training meta-model, etc. Firstly, this paper expounds the concept and framework of algorithm selection based on meta-learning, and reviews related surveys. Secondly, it summarizes the research progress from three aspects: meta-features, meta-learners and meta-model performance measures, introduces typical methods and compares the advantages, disadvantages and application scope of different types of methods. Then, it outlines the application of algorithm selection based on meta-learning in different learning tasks. Next, it utilizes 140 classification datasets, 9 candidate classification algorithms and 5 performance indicators to conduct algorithm selection experiments to compare the performance of different algorithm selection methods. Finally, it analyzes the current challenges and problems, and discusses future development directions.