1 code implementation • 19 May 2018 • Jennifer E. Starling, Jared S. Murray, Carlos M. Carvalho, Radek Bukowski, James G. Scott
This article introduces functional BART, a new approach for functional response regression--that is, estimating a functional mean response f(t) that depends upon a set of scalar covariates x. Functional BART, or funBART, is based on the successful Bayesian Additive Regression Trees (BART) model.
Methodology
no code implementations • 6 Aug 2017 • Wesley Tansey, Jesse Thomason, James G. Scott
We consider the problem of estimating a regression function in the common situation where the number of features is small, where interpretability of the model is a high priority, and where simple linear or additive models fail to provide adequate performance.
no code implementations • 23 Feb 2017 • Wesley Tansey, James G. Scott
We consider the problem of estimating a regression function in the common situation where the number of features is small, where interpretability of the model is a high priority, and where simple linear or additive models fail to provide adequate performance.
1 code implementation • 23 Feb 2017 • Wesley Tansey, Karl Pichotta, James G. Scott
We present an approach to deep estimation of discrete conditional probability distributions.
no code implementations • 1 Dec 2016 • Wesley Tansey, Edward W. Lowe Jr., James G. Scott
Smart phone apps that enable users to easily track their diets have become widespread in the last decade.
no code implementations • 7 Jun 2016 • Wesley Tansey, Karl Pichotta, James G. Scott
CDE Trend Filtering applies a k-th order graph trend filtering penalty to the unnormalized logits of a multinomial classifier network, with each edge in the graph corresponding to a neighboring point on a discretized version of the density.
1 code implementation • 24 May 2015 • Wesley Tansey, James G. Scott
We propose a new algorithm for solving the graph-fused lasso (GFL), a method for parameter estimation that operates under the assumption that the signal tends to be locally constant over a predefined graph structure.
no code implementations • 24 Feb 2015 • Oscar Hernan Madrid Padilla, James G. Scott
We present an approach for penalized tensor decomposition (PTD) that estimates smoothly varying latent factors in multi-way data.
no code implementations • 11 Feb 2015 • Nicholas G. Polson, James G. Scott, Brandon T. Willard
We provide a discussion of convergence of non-descent algorithms with acceleration and for non-convex functions.
1 code implementation • 22 Nov 2014 • Wesley Tansey, Oluwasanmi Koyejo, Russell A. Poldrack, James G. Scott
We also apply the method to a data set from an fMRI experiment on spatial working memory, where it detects patterns that are much more biologically plausible than those detected by standard FDR-controlling methods.
Methodology Applications Computation
2 code implementations • 2 May 2014 • Jesse Windle, Nicholas G. Polson, James G. Scott
Efficiently sampling from the P\'olya-Gamma distribution, ${PG}(b, z)$, is an essential element of P\'olya-Gamma data augmentation.
Computation
no code implementations • 12 Apr 2014 • Mingyuan Zhou, Oscar Hernan Madrid Padilla, James G. Scott
We define a family of probability distributions for random count matrices with a potentially unbounded number of rows and columns.
1 code implementation • 12 Jul 2013 • James G. Scott, Ryan C. Kelly, Matthew A. Smith, Pengcheng Zhou, Robert E. Kass
But this may be inappropriate for many of today's large-scale screening problems, where auxiliary information about each test is often available, and where a combined analysis can lead to poorly calibrated error rates within different subsets of the experiment.
Methodology Applications
no code implementations • 31 May 2013 • James G. Scott, Liang Sun
We present a family of expectation-maximization (EM) algorithms for binary and negative-binomial logistic regression, drawing a sharp connection with the variational-Bayes algorithm of Jaakkola and Jordan (2000).
2 code implementations • 2 May 2012 • Nicholas G. Polson, James G. Scott, Jesse Windle
We propose a new data-augmentation strategy for fully Bayesian inference in models with binomial likelihoods.