New England Machine Learning Day 2018

Semi-Supervised Learning with Competitive Infection Models

Nir Rosenfeld, Harvard University/Amir Globerson, Tel Aviv University

Multiscale Semi-Markov Dynamics for Intracortical Brain-Computer Interfaces

Breaking the n^{-1/2} barrier for permutation-based ranking models

Cheng Mao, MIT/Jonathan Weed, MIT; Philippe Rigollet, MIT; Ashwin Pananjady UC Berkeley; Martin J. Wainwright, UC Berkeley

Application of Breiman and Cutler’s Random Forest Algorithm for Identification of Mutated Genes Responsible for Drug Resistance in M. Tuberculosis Strains

Uma Girkar, MIT/ Ling TengHarvard Medical School; Dr. Gil Alterovitz, Harvard Medical School

Bayesian Nonparametrics in Julia

An Epidemic Modeling Framework For Hashtag Diffusion on Congressional Twitter Networks

Cantay Caliskan, Boston University/ Dino P. Christenson, Boston University

Inference and Learning in Latent Count Models

Combating Imbalanced Data with Generative Adversarial Networks

Rheeya Uppaal, UMass Amherst

Stochastic dynamics of sensory cortical neurons underlie taste-related decision making

Joint Event Detection and Description in Continuous Video Streams

Huijuan Xu, Boston University/ Boyang Li, Liulishuo Silicon Valley AI; LabVasili Ramanishka, Boston University; Leonid Sigal, University of British Columbia; Kate Saenko, Boston University

When Life Gives you Lemmas, Make an Cross-Document Event Coreference Resolution System

Dissociating Linguistic Form and Meaning with Adversarial-Motivational Training

Alexey Romanov, University of Massachussetts/ Anna R., University of Massachusetts Lowell; Anna R., University of Massachusetts Lowell; David D., University of Massachusetts Lowell

Early Syntactic Bootstrapping in an Incremental Memory-Limited Word Learner

Synthetic and Natural Noise Both Break Neural Machine Translation

Yonatan Belinkov, MIT/ Yonatan Bisk, University of Washington

Unbiased Hamiltonian Monte Carlo with couplings

State Abstractions for Lifelong Reinforcement Learning

David Abel, Brown University/ Dilip Arumugam, Brown University; Lucas Lehnert, Brown University; Michael L. Littman, Brown University

Policy and Value Transfer for Lifelong Reinforcement Learning

A Robust Learning Algorithm for Regression Models Using Distributionally Robust Optimization under the Wasserstein Metric

Ruidi Chen, Boston University/ Ioannis Ch. Paschalidis, Boston University

Generalizing Bottleneck Problems

Limits of Learning to Reduce Incompleteness in Partially Observed Networks

Timothy LaRock, Northeastern University/ Sahely Bhadra, Indian Institute of Technology; Tina Eliassi-Rad, Northeastern University

Distributing Frank-Wolfe via Map-Reduce

Non-Parametric Inference for Gaussian Process

Linfeng Liu, Tufts University/ Liping Liu., Tufts University

On the Sample Complexity of Adversarially Robust Generalization

Optimality of Approximate Inference Algorithms on Stable Instances

Hunter Lang, MIT/ David Sontag, MIT; Aravindan Vijayaraghavan, Northwestern University

Graph Distance from the Topological Perspective of Nonbacktracking Cycles

Correlation-based Time Series Analytics

Ramoza Ahsan, Worcester Polytechnic Institute/ Rodica Neamtu, Worcester Polytechnic Institute; Muzammil Bashir, Worcester Polytechnic Institute; Elke Rundensteiner, Worcester Polytechnic Institute; Garbor Sarkozy, Worcester Polytechnic Institute

Learning Deep Embeddings by Learning to Rank

Learning Disentangled Representations of Texts with Application to Biomedical Abstracts

Sarthak Jain, Northeastern University/ Edward Banner, CCIS, Northeastern University; Jan-Willem van de Meent, Northeastern University; Iain J Marshall, King’s College London; Byron C Wallace, CCIS, Northeastern University

Time Series Analysis via Matrix Estimation

Why did they cite that?

Charles Lovering, Worcester Polytechnic Institute/ Jake Whitehill, WPI

Committee-Based Anomaly Detection with Explanations

Multiagent Norm Identification: A Belief-Theoretic Approach for Automatically Identifying Explicitly Represented Norms from Observation

Vasanth Sarathy, Tufts University/ Matthias Scheutz, Tufts University

Improving Emotion Detection with Sub-clip Classification Boosting

Distributionally Robust Submodular Maximization

Matthew Staib, MIT/ Bryan Wilder, USC; Stefanie Jegelka, MIT

Experimental Design under Bradley Terry Model

Deep Learning for Optimal Filtering

Matt Weiss, Worcester Polytechnic Institute/ Randy C. Paffenroth, Worcester Polytechnic Institute; Joshua R. Uzarski, U.S. Army NSRDEC; Jacob R. Whitehill, Worcester Polytechnic Institute

Separation of time scales and direct computation of weights in deep neural networks

An ADMM-Based Universal Framework for Adversarial Attacks on Deep Neural Networks

Pu Zhao, Northeastern University/ Sijia Liu, IBM research; AIKaidi Xu, Northeastern University; Yanzhi Wang, Northeastern University; Xue Lin, Northeastern University

Improving Shape Deformation in Unsupervised Image-to-Image Translation

Learning in POMDPs with Monte Carlo Tree Search

Sammie Katt, Northeastern University/ Frans A. Oliehoek, University of Liverpool; Christopher Amato, Northeastern University

Learning to Place Objects: A Network-based Approach

Practical Data-Dependent Metric Compression with Provable Guarantees

Tal Wagner, MIT/ Piotr Indyk, MIT; Ilya Razenshteyn, Microsoft Research

Hierarchical Disentangled Representations

On the Direction of Discrimination: An Information-Theoretic Analysis of Disparate Impact in Machine Learning

Hao Wang, Harvard University/ Berk Ustun, Harvard University; Flavio P. Calmon, Harvard University

One-shot Learning and Classification in Kids

ShrinkNets: Learning Network Size while Training

Guillaume Leclerc, MIT/ Raul, Castro MIT; Manasi Vartak, MIT; Sam Madden, MIT; Tim Kraska, MIT

Low Variance Gradients for Variational Inference

Topically-Coherent Neural Language Model Conditioned on Arbitrary Features

Xiao Qin, Worcester Polytechnic Institute/ Elke Rundensteiner, Worcester Polytechnic Institute; Xiangnan Kong, Worcester Polytechnic Institute

Inferring Electrification in Developing Nations via Hierarchical Beta Models of Multimodal Observations

Automated software vulnerability detection using Long Short-Term Memory