Unifying Themes in Empirical and Explanation-Based Learning
暂无分享,去创建一个
Publisher Summary This chapter discusses induction and explanation-based learning. A central activity of science is the search for unifying principles that account for apparently diverse phenomena within a single framework. However, recent work in machine learning has tended to emphasize the differences between learning methods. Significant differences certainly exist between explanation-based and empirical methods, but the perceived chasm is far greater than the actual one. This perception has resulted partly from a literature that abounds with rhetorical statements claiming superiority of one method over another. Other causes for the perceived distinction include divergent notations and different measures of performance, which hide the underlying similarity of mechanisms and tasks.
[1] Tom M. Mitchell,et al. Generalization as Search , 2002 .
[2] Michael John Pazzani. Learning causal relationships: an integration of empirical and explanation-based learning methods , 1988 .
[3] J. Gerard Wolff,et al. Language acquisition, data compression and generalization , 1982 .
[4] Steven Minton,et al. Quantitative Results Concerning the Utility of Explanation-based Learning , 1988, Artif. Intell..