• Publications
  • Influence
Variational Autoencoders for Collaborative Filtering
TLDR
A generative model with multinomial likelihood and use Bayesian inference for parameter estimation is introduced and the pros and cons of employing a principledBayesian inference approach are identified and characterize settings where it provides the most significant improvements.
Probability Product Kernels
TLDR
The advantages of discriminative learning algorithms and kernel machines are combined with generative modeling using a novel kernel between distributions to exploit the properties, metrics and invariances of the generative models the authors infer from each datum.
Bayesian face recognition
Graph construction and b-matching for semi-supervised learning
TLDR
Experimental results on both artificial data and real benchmark datasets indicate that b-matching produces more robust graphs and therefore provides significantly better prediction accuracy without any significant change in computation time.
Maximum Entropy Discrimination
TLDR
A general framework for discriminative estimation based on the maximum entropy principle and its extensions is presented and preliminary experimental results are indicative of the potential in these techniques.
Computational Social Science
TLDR
It is found that origination rates of marine bivalves increased significantly almost everywhere immediately after the K-Pg mass extinction event, and a distinct pulse of bivalve diversification in the early Cenozoic was concentrated mainly in tropical and subtropical regions.
Bhattacharyya and Expected Likelihood Kernels
TLDR
A new class of kernels between distributions that permits discriminative estimation via, for instance, support vector machines, while exploiting the properties, assumptions, and invariances inherent in the choice of generative model are introduced.
A Kernel Between Sets of Vectors
TLDR
The kernel between examples is defined as Bhattacharyya's measure of affinity between Gaussians, which is computable in closed form and enjoys many favorable properties, including graceful behavior under transformations, potentially justifying the vector set representation even in cases when more conventional representations also exist.
Graph transduction via alternating minimization
TLDR
This paper introduces a propagation algorithm that more reliably minimizes a cost function over both a function on the graph and a binary label matrix and achieves substantial improvement in accuracy compared to state of the art semi-supervised methods.
Spectral Clustering and Embedding with Hidden Markov Models
TLDR
This article shows that using probabilistic pairwise kernel estimates between parametric models provides improved experimental results for unsupervised clustering and visualization of real and synthetic datasets.
...
...