Christopher DuBois

Stochastic collapsed variational Bayesian inference for latent Dirichlet allocation

There has been an explosion in the amount of digital text information available in recent years, leading to challenges of scale for traditional inference algorithms for topic models. Recent advances in stochastic variational inference algorithms for latent Dirichlet allocation (LDA) have made it feasible to learn topic models on very large-scale corpora, but these methods do not currently take full advantage of the collapsed representation of the model. We propose a stochastic algorithm for collapsed variational Bayesian inference for LDA, which is simpler and more efficient than the state of the art method. In experiments on large-scale text corpora, the algorithm was found to converge faster and often to a better solution than previous methods. Human-subject experiments also demonstrated that the method can learn coherent topics in seconds on small corpora, facilitating the use of topic models in interactive document analysis software.

Modeling relational events via latent classes

Many social networks can be characterized by a sequence of dyadic interactions between individuals. Techniques for analyzing such events are of increasing interest. In this paper, we describe a generative model for dyadic events, where each event arises from one of C latent classes, and the properties of the event (sender, recipient, and type) are chosen from distributions over these entities conditioned on the chosen class. We present two algorithms for inference in this model: an expectation-maximization algorithm as well as a Markov chain Monte Carlo procedure based on collapsed Gibbs sampling. To analyze the models predictive accuracy, the algorithms are applied to multiple real-world data sets involving email communication, international political events, and animal behavior data.

Large capture-range of a single-beam gradient optical trap

We have observed long-range trapping with a single-beam gradient force optical trap. 6 to 10 microm polystyrene beads that are initially approximately 100 microm away from the trap-center can be pulled into the trap-center. Particle-tracking enables us to determine the trajectory of a bead as it moves towards the trap-center and map out a capture zone inside which trapping can occur.

A Multiple Time-scale Computational Model of a Tumor and Its Micro Environment

Experimental evidence suggests that a tumors environment may be critical to designing successful therapeutic protocols: Modeling interactions between a tumor and its environment could improve our understanding of tumor growth and inform approaches to treatment. This paper describes an efficient, flexible, hybrid cellular automaton-based implementation of numerical solutions to multiple time-scale reaction-diffusion equations, applied to a model of tumor proliferation. The growth and maintenance of cells in our simulation depend on the rate of cellular energy (ATP) metabolized from nearby nutrients such as glucose and oxygen. Nutrient consumption rates are functions of local pH as well as local concentrations of oxygen and other fuels. The diffusion of these nutrients is modeled using a novel variation of random-walk techniques. Furthermore, we detail the effects of three boundary update rules on simulations, describing their effects on computational efficiency and biological realism. Qualitative and quantitative results from simulations provide insight on how tumor growth is affected by various environmental changes such as micro-vessel density or lower pH, both of high interest in current cancer research.

Modeling individual email patterns over time with latent variable models

As digital communication devices play an increasingly prominent role in our daily lives, the ability to analyze and understand our communication patterns becomes more important. In this paper, we investigate a latent variable modeling approach for extracting information from individual email histories, focusing in particular on understanding how an individual communicates over time with recipients in their social network. The proposed model consists of latent groups of recipients, each of which is associated with a piecewise-constant Poisson rate over time. Inference of group memberships, temporal changepoints, and rate parameters is carried out via Markov Chain Monte Carlo (MCMC) methods. We illustrate the utility of the model by applying it to both simulated and real-world email data sets.