The Dimension of Self-Directed Learning

被引:0
作者
Devulapalli, Pramith [1 ]
Hanneke, Steve [1 ]
机构
[1] Purdue Univ, Dept Comp Sci, W Lafayette, IN 47907 USA
来源
INTERNATIONAL CONFERENCE ON ALGORITHMIC LEARNING THEORY, VOL 237 | 2024年 / 237卷
关键词
Self-Directed Learning; Online Learning; Mistake-Bounds; QUERIES;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Understanding the self-directed learning complexity has been an important problem that has captured the attention of the online learning theory community since the early 1990s. Within this framework, the learner is allowed to adaptively choose its next data point in making predictions unlike the setting in adversarial online learning. In this paper, we study the self-directed learning complexity in both the binary and multi-class settings, and we develop a dimension, namely SDdim, that exactly characterizes the self-directed learning mistake-bound for any concept class. The intuition behind SDdim can be understood as a two-player game called the "labelling game". Armed with this two-player game, we calculate SDdim on a whole host of examples with notable results on axis-aligned rectangles, VC dimension 1 classes, and linear separators. We demonstrate several learnability gaps with a central focus between self-directed learning and offline sequence learning models that include either the best or worst ordering. Finally, we extend our analysis to the self-directed binary agnostic setting where we derive upper and lower bounds.
引用
收藏
页数:30
相关论文
共 20 条
[1]   Adversarial Laws of Large Numbers and Optimal Regret in Online Classification [J].
Alon, Noga ;
Ben-Eliezer, Omri ;
Dagan, Yuval ;
Moran, Shay ;
Naor, Moni ;
Yogev, Eylon .
STOC '21: PROCEEDINGS OF THE 53RD ANNUAL ACM SIGACT SYMPOSIUM ON THEORY OF COMPUTING, 2021, :447-455
[2]   Queries revisited [J].
Angluin, D .
THEORETICAL COMPUTER SCIENCE, 2004, 313 (02) :175-194
[3]  
Angluin D., 1988, Machine Learning, V2, P319, DOI 10.1007/BF00116828
[4]  
Ben-David S, 2015, Arxiv, DOI arXiv:1507.05307
[5]  
Ben -David Shai, 2009, COLT, V3, P1
[6]   CHARACTERIZATIONS OF LEARNABILITY FOR CLASSES OF (O,...,N)-VALUED FUNCTIONS [J].
BENDAVID, S ;
CESABIANCHI, N ;
HAUSSLER, D ;
LONG, PM .
JOURNAL OF COMPUTER AND SYSTEM SCIENCES, 1995, 50 (01) :74-86
[7]   Online learning versus offline learning [J].
BenDavid, S ;
Kushilevitz, E ;
Mansour, Y .
MACHINE LEARNING, 1997, 29 (01) :45-63
[8]  
Cesa-Bianchi N., 2006, PREDICTION LEARNING, DOI 10.1017/CBO9780511546921
[9]   How to use expert advice [J].
CesaBianchi, N ;
Freund, Y ;
Haussler, D ;
Helmbold, DP ;
Schapire, RE ;
Warmuth, MK .
JOURNAL OF THE ACM, 1997, 44 (03) :427-485
[10]  
Chase H., 2020, P 33 C LEARN THEOR