Online convex combination of ranking models
Open Access
- 6 November 2021
- journal article
- research article
- Published by Springer Nature in User Modeling and User-Adapted Interaction
- Vol. 32 (4) , 649-683
Abstract
As a task of high importance for recommender systems, we consider the problem of learning the convex combination of ranking algorithms by online machine learning. First, we propose a stochastic optimization algorithm that uses finite differences. Our new algorithm achieves close to optimal empirical performance for two base rankers, while scaling well with an increased number of models. In our experiments with five real-world recommendation data sets, we show that the combination offers significant improvement over previously known stochastic optimization techniques. The proposed algorithm is the first effective stochastic optimization method for combining ranked recommendation lists by online machine learning. Secondly, we propose an exponentially weighted algorithm based on a grid over the space of combination weights. We show that the algorithm has near-optimal worst-case performance bound. The bound provides the first theoretical guarantee for non-convex bandits using limited number of evaluations under very general conditions.Keywords
Funding Information
- Hungarian Ministry of Innovation and Technology NRDI Office within the framework of the Hungarian Artificial Intelligence National Laboratory Program
- ELKH Institute for Computer Science and Control
This publication has 34 references indexed in Scilit:
- Image-Based Recommendations on Styles and Substitutes2015
- Temporal influence over the Last.fm social networkSocial Network Analysis and Mining, 2015
- An Empirical Comparison of CMA-ES in Dynamic Environments2012
- Online Learning in Adversarial Lipschitz EnvironmentsLecture Notes in Computer Science, 2010
- Factorization meets the neighborhood2008
- A computational efficient covariance matrix update and a (1+1)-CMA for evolution strategies2006
- Universal parameter optimisation in games based on SPSAMachine Learning, 2006
- The Nonstochastic Multiarmed Bandit ProblemSIAM Journal on Computing, 2002
- Item-based collaborative filtering recommendation algorithms2001
- Multivariate stochastic approximation using a simultaneous perturbation gradient approximationIEEE Transactions on Automatic Control, 1992