Abstract
This paper aims to shed light on achievable limits in active learning. Using minimax analysis techniques, we study the achievable rates of classification error convergence for broad classes of distributions characterized by decision boundary regularity and noise conditions. The results clearly indicate the conditions under which one can expect significant gains through active learning. Furthermore we show that the learning rates derived are tight for “boundary fragment” classes in d-dimensional feature spaces when the feature marginal density is bounded from above and below.
Access provided by Autonomous University of Puebla. Download to read the full chapter text
Chapter PDF
Similar content being viewed by others
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
Mackay, D.J.C.: Information-based objective functions for active data selection. Neural Computation 4, 698–714 (1991)
Cohn, D., Ghahramani, Z., Jordan, M.: Active learning with statistical models. Journal of Artificial Intelligence Research, pp. 129–145 ( 1996)
Freund, Y., Seung, H.S., Shamir, E., Tishby, N.: Selective sampling using the query by committee algorithm. Machine Learning 28(2-3), 133–168 (1997)
Cesa-Bianchi, N., Conconi, A., Gentile, C.: Learning probabilistic linear-threshold classifiers via selective sampling. In: Schölkopf, B., Warmuth, M.K. (eds.) COLT/Kernel 2003. LNCS (LNAI), vol. 2777, Springer, Heidelberg (2003)
Blanchard, G., Geman, D.: Hierarchical testing designs for pattern recognition. The Annals of Statistics 33(3), 1155–1202 (2005)
Dasgupta, S., Kalai, A., Monteleoni, C.: Analysis of perceptron-based active learning. In: Auer, P., Meir, R. (eds.) COLT 2005. LNCS (LNAI), vol. 3559, Springer, Heidelberg (2005)
Dasgupta, S.: Coarse sample complexity bounds for active learning. In: Advances in Neural Information Processing (NIPS) (2005)
Dasgupta, S.: Analysis of a greedy active learning strategy. In: Advances in Neural Information Processing (NIPS) (2004)
Balcan, N., Beygelzimer, A., Langford, J.: Agostic active learning. In: 23rd International Conference on Machine Learning, Pittsburgh, PA, USA (2006)
Castro, R., Willett, R., Nowak, R.: Faster rates in regression via active learning. In: Proceedings of Neural Information Processing Systems (NIPS), extended version (2005), available at http://homepages.cae.wisc.edu/~rcastro/ECE-05-3.pdf
Kääriäinen, M.: On active learning in the non-realizable case. NIPS Workshop on Foundations of Active Learning (2005)
Burnashev, M.V., Zigangirov, K.S.: An interval estimation problem for controlled observations. Problems in Information Transmission 10, 223–231 (1974) (Translated from Problemy Peredachi Informatsii, 10(3),51–61, July-September, 1974). Original article submitted (June 25, 1973)
Hall, P., Molchanov, I.: Sequential methods for design-adaptive estimation of discontinuities in regression curves and surfaces. The Annals of Statistics 31(3), 921–941 (2003)
Golubev, G., Levit, B.: Sequential recovery of analytic periodic edges in the binary image models. Mathematical Methods of Statistics 12, 95–115 (2003)
Bryan, B., Schneider, J., Nichol, R.C., Miller, C.J., Genovese, C.R., Wasserman, L.: Active learning for identifying function threshold boundaries. In: Advances in Neural Information Processing (NIPS) (2005)
Tsybakov, A.: Optimal aggregation of classifiers in statistical learning. The Annals of Statistics 32(1), 135–166 (2004)
Cavalier, L.: Nonparametric estimation of regression level sets. Statistics 29, 131–160 (1997)
Tsybakov, A.B.: On nonparametric estimation of density level sets. The Annals of Statistics 25, 948–969 (1997)
Tsybakov, A.B.: Introduction à l’estimation non-paramétrique. In: Mathématiques et Applications, vol. 41, Springer, Heidelberg (2004)
Korostelev, A.P.: On minimax rates of convergence in image models under sequential design. Statistics & Probability Letters 43, 369–375 (1999)
Korostelev, A., Kim, J.C.: Rates of convergence for the sup-norm risk in image models under sequential designs. Statistics & probability Letters 46, 391–399 (2000)
Castro, R., Nowak, R.: Upper and lower bounds for active learning. In: 44th Annual Allerton Conference on Communication, Control and Computing (2006)
Castro, R.M., Nowak, R.D.: Minimax bounds for active learning. Technical report, ECE Dept. University of Wisconsin - Madison (2007), available at http://homepages.cae.wisc.edu/~rcastro/ECE-07-3.pdf
de Boor, C.: The error in polynomial tensor-product and chung-yao, interpolation. In: LeMéhauté, A., Rabut, C., Schumaker, L., (eds.): Surface Fitting and Multiresolution Methods, Vanderbilt University Press, 35–50 (1997)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2007 Springer Berlin Heidelberg
About this paper
Cite this paper
Castro, R.M., Nowak, R.D. (2007). Minimax Bounds for Active Learning. In: Bshouty, N.H., Gentile, C. (eds) Learning Theory. COLT 2007. Lecture Notes in Computer Science(), vol 4539. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-72927-3_3
Download citation
DOI: https://doi.org/10.1007/978-3-540-72927-3_3
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-72925-9
Online ISBN: 978-3-540-72927-3
eBook Packages: Computer ScienceComputer Science (R0)