no code implementations • 2 Sep 2022 • Han Wu, Jie Yin, Bala Rajaratnam, Jianyuan Guo
By jointly capturing three levels of relational information (entity-level, triplet-level and context-level), HiRe can effectively learn and refine the meta representation of few-shot relations, and consequently generalize very well to new unseen relations.
no code implementations • 11 May 2015 • Alfred O. Hero, Bala Rajaratnam
Sampling regimes can be divided into several categories: 1) the classical asymptotic regime where the variable dimension is fixed and the sample size goes to infinity; 2) the mixed asymptotic regime where both variable dimension and sample size go to infinity at comparable rates; 3) the purely high dimensional asymptotic regime where the variable dimension goes to infinity and the sample size is fixed.
no code implementations • 22 Feb 2015 • Hamed Firouzi, Alfred Hero, Bala Rajaratnam
In the first stage we collect a few ($n$) expensive samples $\{y_i,\mathbf x_i\}_{i=1}^n$, at the full dimension $p\gg n$ of $\mathbf X$, winnowing the number of variables down to a smaller dimension $l < p$ using a type of cross-correlation or regression coefficient screening.
no code implementations • NeurIPS 2014 • Sang-Yun Oh, Onkar Dalal, Kshitij Khare, Bala Rajaratnam
In direct contrast to the parallel work in the Gaussian setting however, this new convex pseudo-likelihood framework has not leveraged the extensive array of methods that have been proposed in the machine learning literature for convex optimization.
no code implementations • 13 May 2014 • Onkar Dalal, Bala Rajaratnam
Over and above estimating a sparse inverse covariance matrix, we also illustrate how to (1) incorporate constraints on the (bivariate) correlations and, (2) incorporate equality (equisparsity) or linear constraints between individual inverse covariance elements.
no code implementations • 9 Oct 2013 • Dhafer Malouche, Bala Rajaratnam, Benjamin T. Rolfs
First, we discuss the pairwise and global Markov properties for undirected and bidirected models, using the pseudographoid and reverse-pseudographoid rules which are weaker conditions than the typically used intersection and composition rules.
no code implementations • 20 Jul 2013 • Kshitij Khare, Sang-Yun Oh, Bala Rajaratnam
As none of the popular methods proposed for solving pseudo-likelihood based objective functions have provable convergence guarantees, it is not clear if corresponding estimators exist or are even computable, or if they actually yield correct partial correlation graphs.
no code implementations • 10 Mar 2013 • Hamed Firouzi, Bala Rajaratnam, Alfred Hero
We introduce a new approach to variable selection, called Predictive Correlation Screening, for predictor design.
1 code implementation • NeurIPS 2012 • Dominique Guillot, Bala Rajaratnam, Benjamin T. Rolfs, Arian Maleki, Ian Wong
In this paper, a proximal gradient method (G-ISTA) for performing L1-regularized covariance matrix estimation is presented.