Learn More
This paper introduces a fast Bayesian online expectation maximiza-tion (BOEM) algorithm for multinomial mixtures. Using some properties of the Dirichlet distribution, we derive expressions for adaptive learning rates that depend solely on the data and the prior's hyper-parameters. As a result, we avoid the problem of having to tune the learning rates using(More)
We present a novel, flexible, statistical approach to modeling music, images and text jointly. The technique is based on multi-modal mixture models and efficient computation using online EM. The learned models can be used to browse mul-timedia databases, to query on a multimedia database using any combination of music, images and text (lyrics and other(More)
We accept this thesis as conforming to the required standard Abstract The EM algorithm is one of the most popular statistical learning algorithms. Unfortunately, it is a batch learning method. For large data sets and real-time systems, we need to develop on-line methods. In this thesis, we present a comprehensive study of on-line EM algorithms. We use(More)
  • 1