Combining Online Learning Guarantees

We show how to take any two parameter-free online learning algorithms with different regret guarantees and obtain a single algorithm whose regret is the minimum of the two base algorithms. Our method is embarrassingly simple: just add the iterates. This trick can generate efficient algorithms that adapt to many norms simultaneously, as well as providing diagonal-style algorithms that still maintain dimension-free guarantees. We then proceed to show how a variant on this idea yields a black-box procedure for generating optimistic online learning algorithms. This yields the first optimistic regret guarantees in the unconstrained setting and generically increases adaptivity. Further, our optimistic algorithms are guaranteed to do no worse than their non-optimistic counterparts regardless of the quality of the optimistic estimates provided to the algorithm.

[1]  Francesco Orabona,et al.  Dimension-Free Exponentiated Gradient , 2013, NIPS.

[2]  Yoram Singer,et al.  Adaptive Subgradient Methods for Online Learning and Stochastic Optimization , 2011, J. Mach. Learn. Res..

[3]  Tatiana Tommasi,et al.  Training Deep Networks without Learning Rates Through Coin Betting , 2017, NIPS.

[4]  Elad Hazan,et al.  Extracting certainty from uncertainty: regret bounded by variation in costs , 2008, Machine Learning.

[5]  Matthew J. Streeter,et al.  Adaptive Bound Optimization for Online Convex Optimization , 2010, COLT 2010.

[6]  Francesco Orabona,et al.  Unconstrained Online Linear Learning in Hilbert Spaces: Minimax Algorithms and Normal Approximations , 2014, COLT.

[7]  H. Brendan McMahan,et al.  A survey of Algorithms and Analysis for Adaptive Online Learning , 2014, J. Mach. Learn. Res..

[8]  Massimiliano Pontil,et al.  Empirical Bernstein Bounds and Sample-Variance Penalization , 2009, COLT.

[9]  John Langford,et al.  Normalized Online Learning , 2013, UAI.

[10]  Karthik Sridharan,et al.  Adaptive Online Learning , 2015, NIPS.

[11]  Mehryar Mohri,et al.  Accelerating Online Convex Optimization via Adaptive Prediction , 2016, AISTATS.

[12]  Claudio Gentile,et al.  On the generalization ability of on-line learning algorithms , 2001, IEEE Transactions on Information Theory.

[13]  Rong Jin,et al.  25th Annual Conference on Learning Theory Online Optimization with Gradual Variations , 2022 .

[14]  Martin Zinkevich,et al.  Online Convex Programming and Generalized Infinitesimal Gradient Ascent , 2003, ICML.

[15]  Shai Shalev-Shwartz,et al.  Online Learning and Online Convex Optimization , 2012, Found. Trends Mach. Learn..

[16]  Francesco Orabona,et al.  Coin Betting and Parameter-Free Online Learning , 2016, NIPS.

[17]  Karthik Sridharan,et al.  Online Learning: Sufficient Statistics and the Burkholder Method , 2018, COLT.

[18]  Karthik Sridharan,et al.  On Equivalence of Martingale Tail Bounds and Deterministic Regret Inequalities , 2015, COLT.

[19]  Francesco Orabona,et al.  Black-Box Reductions for Parameter-free Online Learning in Banach Spaces , 2018, COLT.

[20]  Matthew J. Streeter,et al.  No-Regret Algorithms for Unconstrained Online Convex Optimization , 2012, NIPS.

[21]  Francesco Orabona,et al.  Simultaneous Model Selection and Optimization through Parameter-free Stochastic Learning , 2014, NIPS.

[22]  Karthik Sridharan,et al.  Online Learning with Predictable Sequences , 2012, COLT.

[23]  Mehryar Mohri,et al.  Parameter-Free Online Learning via Model Selection , 2017, NIPS.