His work spans many areas: sublinear algorithms, graph algorithms, graph modeling, and scalable computation for large data sets.High-dimensional random geometric graphs Miklos Racz, Princeton Abstract: I will talk about two natural random geometric graph models, where connections between vertices depend on distances between latent d-dimensional labels.Tags: How To Write A Literature Review ConclusionUnc Admissions EssaysWrite A Reaction PaperSolving Trigonometry ProblemsTheories Of Crime And Deviance EssayCv Personal Statement Sales ManagerTricks To Solve Trigonometry Problems
Bio: Sid Banerjee is an assistant professor in the School of Operations Research and Information Engineering (ORIE) at Cornell.
His research is on stochastic modeling, and the design of algorithms and incentives for large-scale systems.
Where the researcher has knowledge of the data generating process, there are large returns to choosing an appropriate regularization that emphasizes balance on key covariates --- in general, there is little reason to choose the default regularization implied by the original SC approach.
We explore several alternative approaches for setting the regularization, including cross-validation.
In each of these results, there is an interesting interplay of combinatorics, randomized algorithms, and social network analysis. Seshadhri is an assistant professor of Computer Science at the University of California, Santa Cruz.
Prior to joining UCSC, he was a researcher at Sandia National Labs, Livermore in the Information Security Sciences department, during 2010-2014.We are particularly interested in the high-dimensional case when d is large.We study a basic hypothesis testing problem: can we distinguish a random geometric graph from an Erdos-Renyi random graph (which has no geometry)?Third, we conduct extensive simulation studies to assess the performance of these different estimators in practice.Finally, we use these ideas to assess the impact of the 2012 Kansas tax cuts on economic growth, finding persistent negative effects.The key idea is to construct a weighted average of control units, known as a synthetic control (SC), that minimizes imbalance of pre-treatment outcomes between the treated unit and synthetic control.Our main result is that synthetic control weights are numerically equivalent to inverse propensity score weights (IPW) with pre-treatment outcomes as covariates and heavy regularization of the propensity score model.His applied research focuses on working with governments on using data to design, implement, and evaluate policies. Seshadhri, UC Santa Cruz Abstract: As an algorithms researcher, the existence of heuristics for various computational graph problems (like finding cliques) on real-world graphs bothers me to no end.Prior to his doctoral studies, Feller served as Special Assistant to the Director at the White House Office of Management and Budget and worked at the Center on Budget and Policy Priorities. What is it about (say) social networks that make classically hard problems easy?In the second, we show how to estimate the degree distribution of a graph by sampling a tiny portion of it, by exploiting the heavy tailed structure of the degree distribution.This result uses some recent theoretical techniques in sublinear algorithms to simulate uniform random edge queries through uniform random vertices.