Evan Racah
YOU?
Author Swipe
View article: Hierarchical Model-Based Imitation Learning for Planning in Autonomous Driving
Hierarchical Model-Based Imitation Learning for Planning in Autonomous Driving Open
We demonstrate the first large-scale application of model-based generative adversarial imitation learning (MGAIL) to the task of dense urban self-driving. We augment standard MGAIL using a hierarchical model to enable generalization to arb…
View article: Slot Contrastive Networks: A Contrastive Approach for Representing Objects
Slot Contrastive Networks: A Contrastive Approach for Representing Objects Open
Unsupervised extraction of objects from low-level visual data is an important goal for further progress in machine learning. Existing approaches for representing objects without labels use structured generative models with static images. T…
View article: The LoCA Regret: A Consistent Metric to Evaluate Model-Based Behavior in Reinforcement Learning
The LoCA Regret: A Consistent Metric to Evaluate Model-Based Behavior in Reinforcement Learning Open
Deep model-based Reinforcement Learning (RL) has the potential to substantially improve the sample-efficiency of deep RL. While various challenges have long held it back, a number of papers have recently come out reporting success with dee…
View article: Supervise Thyself: Examining Self-Supervised Representations in Interactive Environments
Supervise Thyself: Examining Self-Supervised Representations in Interactive Environments Open
Self-supervised methods, wherein an agent learns representations solely by observing the results of its actions, become crucial in environments which do not provide a dense reward signal or have labels. In most cases, such methods are used…
View article: Supervise Thyself: Examining Self-Supervised Representations in\n Interactive Environments
Supervise Thyself: Examining Self-Supervised Representations in\n Interactive Environments Open
Self-supervised methods, wherein an agent learns representations solely by\nobserving the results of its actions, become crucial in environments which do\nnot provide a dense reward signal or have labels. In most cases, such methods\nare u…
View article: Unsupervised State Representation Learning in Atari
Unsupervised State Representation Learning in Atari Open
State representation learning, or the ability to capture latent generative factors of an environment, is crucial for building intelligent agents that can perform a wide variety of tasks. Learning such representations without supervision fr…
View article: Deep Neural Networks for Physics Analysis on low-level whole-detector data at the LHC
Deep Neural Networks for Physics Analysis on low-level whole-detector data at the LHC Open
There has been considerable recent activity applying deep convolutional neural nets (CNNs) to data from particle physics experiments. Current approaches on ATLAS/CMS have largely focussed on a subset of the calorimeter, and for identifying…
View article: Deep Neural Networks for Physics Analysis on low-level whole-detector\n data at the LHC
Deep Neural Networks for Physics Analysis on low-level whole-detector\n data at the LHC Open
There has been considerable recent activity applying deep convolutional\nneural nets (CNNs) to data from particle physics experiments. Current\napproaches on ATLAS/CMS have largely focussed on a subset of the calorimeter,\nand for identify…
View article: Deep learning with raw data from Daya Bay
Deep learning with raw data from Daya Bay Open
The Daya Bay experiment uses reactor antineutrino disappearance to measure the θ13 neutrino oscillation parameter. In this proceeding, the convolutional autoencoder machine learning technique is tested against a well-understood uncorrelate…
View article: Deep Learning at 15PF: Supervised and Semi-Supervised Classification for Scientific Data
Deep Learning at 15PF: Supervised and Semi-Supervised Classification for Scientific Data Open
This paper presents the first, 15-PetaFLOP Deep Learning system for solving scientific pattern classification problems on contemporary HPC architectures. We develop supervised convolutional architectures for discriminating signals in high-…
View article: Semi-Supervised Detection of Extreme Weather Events in Large Climate Datasets
Semi-Supervised Detection of Extreme Weather Events in Large Climate Datasets Open
The detection and identification of extreme weather events in large scale climate simulations is an important problem for risk management, informing governmental policy decisions and advancing our basic understanding of the climate system.…
View article: ExtremeWeather: A large-scale climate dataset for semi-supervised detection, localization, and understanding of extreme weather events
ExtremeWeather: A large-scale climate dataset for semi-supervised detection, localization, and understanding of extreme weather events Open
Then detection and identification of extreme weather events in large-scale climate simulations is an important problem for risk management, informing governmental policy decisions and advancing our basic understanding of the climate system…
View article: Revealing Fundamental Physics from the Daya Bay Neutrino Experiment Using Deep Neural Networks
Revealing Fundamental Physics from the Daya Bay Neutrino Experiment Using Deep Neural Networks Open
Experiments in particle physics produce enormous quantities of data that must be analyzed and interpreted by teams of physicists. This analysis is often exploratory, where scientists are unable to enumerate the possible types of signal pri…
View article: Matrix Factorizations at Scale: a Comparison of Scientific Data Analytics in Spark and C+MPI Using Three Case Studies
Matrix Factorizations at Scale: a Comparison of Scientific Data Analytics in Spark and C+MPI Using Three Case Studies Open
We explore the trade-offs of performing linear algebra using Apache Spark, compared to traditional C and MPI implementations on HPC platforms. Spark is designed for data analytics on cluster computing platforms with access to local disks a…
View article: Matrix Factorization at Scale: a Comparison of Scientific Data Analytics in Spark and C+MPI Using Three Case Studies
Matrix Factorization at Scale: a Comparison of Scientific Data Analytics in Spark and C+MPI Using Three Case Studies Open
We explore the trade-offs of performing linear algebra using Apache Spark, compared to traditional C and MPI implementations on HPC platforms. Spark is designed for data analytics on cluster computing platforms with access to local disks a…
View article: Application of Deep Convolutional Neural Networks for Detecting Extreme Weather in Climate Datasets
Application of Deep Convolutional Neural Networks for Detecting Extreme Weather in Climate Datasets Open
Detecting extreme events in large datasets is a major challenge in climate science research. Current algorithms for extreme event detection are build upon human expertise in defining events based on subjective thresholds of relevant physic…
View article: PANDA: Extreme Scale Parallel K-Nearest Neighbor on Distributed Architectures
PANDA: Extreme Scale Parallel K-Nearest Neighbor on Distributed Architectures Open
Computing $k$-Nearest Neighbors (KNN) is one of the core kernels used in many\nmachine learning, data mining and scientific computing applications. Although\nkd-tree based $O(\\log n)$ algorithms have been proposed for computing KNN, due\n…
View article: A Multi-Platform Evaluation of the Randomized CX Low-Rank Matrix Factorization in Spark
A Multi-Platform Evaluation of the Randomized CX Low-Rank Matrix Factorization in Spark Open
We investigate the performance and scalability of the randomized CX low-rank matrix factorization and demonstrate its applicability through the analysis of a 1TB mass spectrometry imaging (MSI) dataset, using Apache Spark on an Amazon EC2 …
View article: Revealing Fundamental Physics from the Daya Bay Neutrino Experiment Using Deep Neural Networks
Revealing Fundamental Physics from the Daya Bay Neutrino Experiment Using Deep Neural Networks Open
Experiments in particle physics produce enormous quantities of data that must be analyzed and interpreted by teams of physicists. This analysis is often exploratory, where scientists are unable to enumerate the possible types of signal pri…