arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1812.03472cs.LGstat.ML
收藏

课程学习理论,含凸损失函数

Theory of Curriculum Learning, with Convex Loss Functions

Hebrew University of Jerusalem(耶路撒冷希伯来大学)

Daphna Weinshall, Dan Amir

AI总结:

本研究为课程学习提供理论分析,证明在凸损失下收敛率随难度分数单调变化,并调和了课程学习与困难数据挖掘的矛盾。

摘要:

课程学习——即通过按有意义的顺序(从易到难)逐步向学习者展示示例来进行教学的思想,早在机器学习领域就已被研究。尽管基于这一概念的方法已被经验性地证明能提升多种学习算法的性能,但即便在简单情形下,也缺乏相应的理论分析。为弥补这一不足,我们首先形式化定义了一个理想的难度分数——即最优假设在给定数据点上的损失。我们基于该分数,在两个凸问题中分析了课程学习可能带来的贡献:线性回归,以及通过铰链损失最小化进行的二分类。我们证明,在这两种情形下,期望收敛率随理想难度分数的增加而单调递减,这与先前的经验结果一致。我们还证明,当理想难度分数固定时,收敛率随当前假设在每个点上的损失增加而单调递增。我们讨论了这些结果如何调和两个看似矛盾的启发式方法:一方面是基于课程学习,另一方面是困难数据挖掘。

英文摘要:

Curriculum Learning - the idea of teaching by gradually exposing the learner to examples in a meaningful order, from easy to hard, has been investigated in the context of machine learning long ago. Although methods based on this concept have been empirically shown to improve performance of several learning algorithms, no theoretical analysis has been provided even for simple cases. To address this shortfall, we start by formulating an ideal definition of difficulty score - the loss of the optimal hypothesis at a given datapoint. We analyze the possible contribution of curriculum learning based on this score in two convex problems - linear regression, and binary classification by hinge loss minimization. We show that in both cases, the expected convergence rate decreases monotonically with the ideal difficulty score, in accordance with earlier empirical results. We also prove that when the ideal difficulty score is fixed, the convergence rate is monotonically increasing with respect to the loss of the current hypothesis at each point. We discuss how these results bring to term two apparently contradicting heuristics: curriculum learning on the one hand, and hard data mining on the other.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑