Pairwise Comparisons with Flexible Time-Dynamics
| Autor: | Victor Kristof, Matthias Grossglauser, Lucas Maystre |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2019 |
| Předmět: |
FOS: Computer and information sciences
Computer Science - Machine Learning Covariance function Computer science Posterior probability Bayesian probability Inference Machine Learning (stat.ML) 02 engineering and technology pairwise comparisons Machine learning computer.software_genre Bayesian inference Machine Learning (cs.LG) symbols.namesake models ranking Statistics - Machine Learning 020204 information systems 0202 electrical engineering electronic engineering information engineering Gaussian process games bayesian inference business.industry Statistical model Kalman filter symbols 020201 artificial intelligence & image processing Pairwise comparison kalman filter Artificial intelligence time series sports business computer |
| Zdroj: | KDD |
| Popis: | Inspired by applications in sports where the skill of players or teams competing against each other varies over time, we propose a probabilistic model of pairwise-comparison outcomes that can capture a wide range of time dynamics. We achieve this by replacing the static parameters of a class of popular pairwise-comparison models by continuous-time Gaussian processes; the covariance function of these processes enables expressive dynamics. We develop an efficient inference algorithm that computes an approximate Bayesian posterior distribution. Despite the flexbility of our model, our inference algorithm requires only a few linear-time iterations over the data and can take advantage of modern multiprocessor computer architectures. We apply our model to several historical databases of sports outcomes and find that our approach outperforms competing approaches in terms of predictive performance, scales to millions of observations, and generates compelling visualizations that help in understanding and interpreting the data. Accepted at KDD 2019 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|