Research

  • Impact Areas
  • Research Areas

Algorithms & Theory

Foundational work that includes complexity, parallel computing and game theory

AI & ML

Spanning natural language processing, deep-learning, computer vision and more

Computational Biology

Understanding disease via epigenomics, gene regulation and bioinformatics

Computer Architecture

How CPUs, memory and other systems are designed and organized

Graphics & Vision

Teaching computers to interpret, create and animate visual data

Human-Computer Interaction

Software and hardware that let us naturally interact with technology

Programming Languages & Software Engineering

From compilers and verification to software design and engineering

Robotics

Vision, actuation, sensing and manipulation of machines

Security & Cryptography

Developing technologies to prevent and recover from cyber-attacks

Systems & Networking

From distributed systems and databases to wireless

Filter options
Filter options

Filter options Close

Role

Media

Research Areas

Impact Areas

Reset Filters

1 People Results matching all criteria

Gregory Falco

Gregory Falco

Research Affiliate
Dr. Falco is a hacker and expert in IoT and IIoT cybersecurity.
Gregory Falco
Dr. Falco is a hacker and expert in IoT and IIoT cybersecurity.

9 Group Results

9 Project Results

Project

CliNER: Clinical Concept Extraction

Cognitive AI Community of Research
Applied Machine Learning Community of Research
Clinical concept extraction (CCE) of named entities - such as problems, tests, and treatments - aids in forming an understanding of notes and provides a foundation for many downstream clinical decision-making tasks. Historically, this task has been posed as a standard named entity recognition (NER) sequence tagging problem, and solved with feature-based methods using hand-engineered domain knowledge. Recent advances, however, have demonstrated the efficacy of LSTM-based models for NER tasks, including CCE. This work presents CliNER 2.0, a simple-to-install, open-source tool for extracting concepts from clinical text. CliNER 2.0 uses a word- and character- level LSTM model, and achieves state-of-the-art performance. For ease of use, the tool also includes pre-trained models available for public use.

Leads

Willie

William Boag

Szolovits

Peter Szolovits

default headshot

Anna Rumshisky

Research Areas

AI & ML

Impact Areas

Energy

Project

PrivacyML - A Privacy Preserving Framework for Machine Learning

Decentralized Information Group
Cognitive AI Community of Research
Computing & Society Community of Research
We are developing a general framework that enforces privacy transparently enabling different kinds of machine learning to be developed that are automatically privacy preserving.
default headshot
default headshot
Kagal-headshot

Lalana Kagal

Vaikkunth

Vaikkunth Mugunthan

Leads

Kagal-headshot

Lalana Kagal

Vaikkunth

Vaikkunth Mugunthan

Research Areas

AI & ML
Security & Cryptography

Impact Areas

Big Data
Cybersecurity
Education
Energy
Entertainment
Health Care
Internet of Things
Manufacturing
Transportation
default headshot
default headshot
 
Kagal-headshot

Lalana Kagal

Vaikkunth

Vaikkunth Mugunthan

17 People Results

Headshot - Dr. R. David Edelman

R. David Edelman

Director, Project on Technology, Economy & National Security
Machine learning, cybersecurity, governance
Headshot - Dr. R. David Edelman
T 258-6389
Machine learning, cybersecurity, governance
Gregory Falco

Gregory Falco

Research Affiliate
Dr. Falco is a hacker and expert in IoT and IIoT cybersecurity.
Gregory Falco
Dr. Falco is a hacker and expert in IoT and IIoT cybersecurity.
Lozano-Perez

PI
Core/Dual

Tomas Lozano-Perez

Professor
Robotics, artificial intelligence
Lozano-Perez
T 253-7889
RM 32-333

Lead

Learning and Intelligent Systems
Learning and Optimization
Robotic Planning and Manipulation with Force Information
Reliable Robotic Construction of Multi-object Arrangements
Learning and Planning with Compressed Linear Expectation Models
Active Learning of Models for Planning
Task and Motion Planning for Autonomous Robots
Robust Intelligent Robots

Member

Robust Intelligent Robots
Embodied Intelligence Community of Research
Robotics, artificial intelligence
Rus

PI
Core/Dual

Daniela Rus

Director
CSAIL director, robotics, AI, data science, privacy
Rus
T 258-7567
RM 32/ 32-368/ 374

Lead

Distributed Robotics Laboratory
Social Network Extraction from GPS Datasets with Coresets
Safety Standards for Autonomous Vehicles
Soft Contact Modeling
Practical Secure Computation
Self-Driving Vehicles
Robust Learning for Autonomous Vehicles
Brain-Controlled Robots
Planning in Congestion for Autonomous Vehicles
Socially-Compliant Behavior for Autonomous Driving
Coresets for Machine Learning Algorithms
M-Blocks Modular Robotics
Robot Compiler
Distributed Robot Garden
Route and Motion Planning for Mobility-on-Demand with Rules of the Road

Member

Social Network Extraction from GPS Datasets with Coresets
Task-Driven Co-Optimization Of Articulated Robots
Wearable Blind Navigation
Drones that Drive
Better Models for Ride-Sharing
SoFi - The Soft Robotic Fish
Controlling drones and other robots with gestures
Using Muscle Signals to Lift Objects with Robots
Embodied Intelligence Community of Research
Helping robots learn using demonstrations
CSAIL director, robotics, AI, data science, privacy
JShah-headshot

PI
Core/Dual

Julie Shah

Associate Professor
Robotics, machine learning, artificial intelligence
JShah-headshot
T 324-4879

Lead

Interactive Robotics Group
Robots that collaborate with humans to form team plans
Robots that Quickly Adjust Team Plans to Disturbance
Human-Guided Transfer for Robot Learning
Efficient Communication for Human-Machine Teams

Member

Apprenticeship Scheduling: Learning to Schedule From Human Experts
Embodied Intelligence Community of Research
Robotics, machine learning, artificial intelligence
solomon-headshot

PI
Core/Dual

Justin Solomon

Assistant Professor
Geometry, graphics, vision, machine learning, optimization.
solomon-headshot
RM 32-D460

Lead

Computer Graphics Group
Geometric Data Processing Group
Geometry in Large-Scale Machine Learning
Political Geometry: Establishing Fair Mathematical Standards for Political Redistricting
Sensible Deep Learning for 3D Data
Optimal transport for statistics and machine learning
Geometry and topology for scientific computing and shape analysis

Member

Visual Computing Community of Research
Geometry, graphics, vision, machine learning, optimization.
Williams

PI
Core/Dual

Brian Williams

Professor
Robotics, autonomy, artificial intelligence, machine learning
Williams
T 253-2739
RM 32-227

Lead

Model-Based Embedded and Robotics Systems Group
RADMAX
Geordi: A Driver's Assistant for Risk-bounded Maneuvering
Robots Adapting to Human Intent
Risk-aware Network Resource Allocation
Uhura: Personal Assistant that Manages Risk
Artificial Social Intelligence for Successful Teams (ASIST)

Member

Cognitive AI Community of Research
Robotics, autonomy, artificial intelligence, machine learning

18 News Results

Researchers at MIT’s Microsystems Technology Laboratories have built a low-power chip specialized for automatic speech recognition. With power savings of 90 to 99 percent, it could make voice control practical for relatively simple electronic devices.
AI & ML Human-Computer Interaction Energy Internet of Things Manufacturing

Voice control everywhere

The butt of jokes as little as 10 years ago, automatic speech recognition is now on the verge of becoming people’s chief means of interacting with their principal computing devices. In anticipation of the age of voice-controlled electronics, MIT researchers have built a low-power chip specialized for automatic speech recognition. Whereas a cellphone running speech-recognition software might require about 1 watt of power, the new chip requires between 0.2 and 10 milliwatts, depending on the number of words it has to recognize.

Voice control everywhere

Image: Jose-Luis Olivares
Algorithms & Theory AI & ML Systems & Networking Big Data Energy

What better wind-speed prediction can do for the energy industry

When a power company wants to build a new wind farm, it generally hires a consultant to make wind speed measurements at the proposed site for eight to 12 months. Those measurements are correlated with historical data and used to assess the site’s power-generation capacity.This month CSAIL researchers will present a new statistical technique that yields better wind-speed predictions than existing techniques do — even when it uses only three months’ worth of data. That could save power companies time and money, particularly in the evaluation of sites for offshore wind farms, where maintaining measurement stations is particularly costly.

What better wind-speed prediction can do for the energy industry