Budgeted Prediction with Expert Advice

We consider a budgeted variant of the problem of learning from expert advice with N experts. Each queried expert incurs a cost and there is a given budget B on the total cost of experts that can be queried in any prediction round. We provide an online learning algorithm for this setting with regret after T prediction rounds bounded by O(√C/B log(N)T, where C is the total cost of all experts. We complement this upper bound with a nearly matching lower bound Ω (√C/B T) on the regret of any algorithm for this problem. We also provide experimental validation of our algorithm.

[1]  David Haussler,et al.  How to use expert advice , 1993, STOC.

[2]  Shai Ben-David,et al.  Learning with restricted focus of attention , 1993, COLT '93.

[3]  Peter Auer,et al.  The Nonstochastic Multiarmed Bandit Problem , 2002, SIAM J. Comput..

[4]  Vijay V. Vazirani,et al.  Approximation Algorithms , 2001, Springer Berlin Heidelberg.

[5]  Horst Bischof,et al.  Real-Time Tracking via On-line Boosting , 2006, BMVC.

[6]  Ohad Shamir,et al.  Efficient Learning with Partially Observed Attributes , 2010, ICML.

[7]  Thierry Bertin-Mahieux,et al.  The Million Song Dataset , 2011, ISMIR.

[8]  Archie C. Chapman,et al.  Knapsack Based Optimal Policies for Budget-Limited Multi-Armed Bandits , 2012, AAAI.

[9]  Sanjeev Arora,et al.  The Multiplicative Weights Update Method: a Meta-Algorithm and Applications , 2012, Theory Comput..

[10]  Elad Hazan,et al.  Linear Regression with Limited Observation , 2012, ICML.

[11]  Aleksandrs Slivkins,et al.  Bandits with Knapsacks , 2013, 2013 IEEE 54th Annual Symposium on Foundations of Computer Science.

[12]  Noga Alon,et al.  From Bandits to Experts: A Tale of Domination and Independence , 2013, NIPS.

[13]  Yi Wu,et al.  Online Object Tracking: A Benchmark , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[14]  Russell Greiner,et al.  Online Learning with Costly Features and Labels , 2013, NIPS.

[15]  Satyen Kale,et al.  Multiarmed Bandits With Limited Expert Advice , 2013, COLT.

[16]  Koby Crammer,et al.  Prediction with Limited Advice and Multiarmed Bandits with Paid Observations , 2014, ICML.