Synonyms
Definition
Given a set of unlabeled training examples, an active learning algorithm is allowed to select a number of training examples (one by one or in batches) and to obtain their target value at a certain cost per example. Next, the learning algorithm should predict the target value of a number of unseen test examples, incurring a certain cost per mistake. The goal of the active learning algorithm is to minimize the total cost.
Characteristics
Obtaining target values of training examples may be expensive. For example, in experimental research requiring microarrays, bioassays, patients, clinical trials, mass-spectrography, crystallography, or other physical experiments in order to collect data, the amount of available resources limits the amount of experimental data which can be obtained.
In an active learning setting, the learning algorithm must take this economic reality into account. Every target value has a cost, and once a model...
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Angluin D (1988) Queries and concept-learning. Mach Learn 2:319–342
Cano-Odena A, Spilliers M, Dedroog T, De Grave K, Ramon J, Vankelecom IFJ (2010) Micropollutant removal via genetic algorithms and high throughput experimentation. J Membr Sci 366(12):25–32
De Grave K, Ramon J, De Raedt L (2008) Active learning for high throughput screening. In: Proceedings of the eleventh international conference on discovery science, Budapest. Lecture Notes in Computer Science, vol 5255, pp 185–196
Guestrin C, Krause A, Singh AP (2005) Near-optimal sensor placement in Gaussian processes. In: Proceedings of the 22nd international conference on machine learning, Bonn, pp 265–272
Kearns M, Vazzirani U (1994) An introduction to computational learning theory. MIT Press, Cambridge, MA
King RD, Whelan KE, Jones FM, Reiser PG, Bryant CH, Muggleton SH, Kell DB, Oliver SG (2004) Functional genomic hypothesis generation and experimentation by a robot scientist. Nature 427:247–252
Liere R, Tadepalli P (1997) Active learning with committees for text categorization. In: Proceedings of the 14th conference of the American association for artificial intelligence (AAAI-97), Providence, pp 591–596
Lizotte D, Wang T, Bowling M, Schuurmans D (2007) Automatic gait optimization with Gaussian process regression. In: Proceedings of the 20th international joint conference on artificial intelligence, Hyderabad, pp 944–949
Tong S, Koller D (2001) Active learning for structure in Bayesian networks. In: Proceedings of the seventeenth International joint conference on artificial intelligence (IJCAI), Seattle. Morgan Kaufman, Washington, pp 863–869
Tong S, Koller D (2001b) Support vector machine active learning with applications to text classification. J Mach Learn Res 2:45–66
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer Science+Business Media, LLC
About this entry
Cite this entry
Ramon, J. (2013). Active Learning. In: Dubitzky, W., Wolkenhauer, O., Cho, KH., Yokota, H. (eds) Encyclopedia of Systems Biology. Springer, New York, NY. https://doi.org/10.1007/978-1-4419-9863-7_605
Download citation
DOI: https://doi.org/10.1007/978-1-4419-9863-7_605
Publisher Name: Springer, New York, NY
Print ISBN: 978-1-4419-9862-0
Online ISBN: 978-1-4419-9863-7
eBook Packages: Biomedical and Life SciencesReference Module Biomedical and Life Sciences