Discusses the Dirichlet distribution, Bayesian inference, posterior mean and variance, conjugate priors, and predictive distribution in the Dirichlet-Multinomial model.
Delves into the fundamental limits of gradient-based learning on neural networks, covering topics such as binomial theorem, exponential series, and moment-generating functions.
Explores the Bayesian extension of HMM for robot action segmentation and modeling, limitations of classical HMMs, and motion capture data segmentation.
Explores learning latent models in graphical structures, focusing on scenarios with incomplete samples and introducing the notion of distance among variables.
Covers topic models, focusing on Latent Dirichlet Allocation, clustering, GMMs, Dirichlet distribution, LDA learning, and applications in digital humanities.
