Research

MIT researchers have employed a new machine-learning technique to substantially reduce false positives in fraud-detecting technologies.

Reducing false positives in credit card fraud detection

Model extracts granular behavioral patterns from transaction data to more accurately flag suspicious activity.

September 20, 2018

MIT computer scientists have developed a system that learns to identify objects within an image, based on a spoken description of the image.

Machine-learning system tackles speech and object recognition, all at once

Model learns to pick out objects within an image, using spoken descriptions.

September 18, 2018

Four GapR subunits assemble as a dimer-of-dimer complex, forming a clamp around overtwisted DNA, which arises during DNA replication and transcription.

Detangling DNA replication

Researchers identify an essential protein that helps enzymes relax overtwisted DNA so each strand can be copied during cell division.

September 17, 2018

It's often unclear how to turn molecular knowledge into new therapeutics, especially for cancers with no clear genetic cause like medulloblastoma.

Protein analysis uncovers new medulloblastoma subtypes

Research team shows the power of proteomics to discover new drug targets and help develop therapeutic strategies with fewer long-term side effects.

September 14, 2018

Edward Doddridge, a postdoc in the Department of Earth, Atmospheric and Planetary Sciences at MIT, co-developed a model to investigate the mechanism behind phytoplankton growth within subtropical gyres.

How phytoplankton survive in ocean gyres with low nutrient supplies

Model of nutrient recycling may explain longstanding mystery.

September 14, 2018

MIT researchers have developed a module that helps artificial-intelligence systems fill in the gaps between video frames to improve activity recognition.

Helping computers fill in the gaps between video frames

Machine learning system efficiently recognizes activities by observing how objects change in only a few key frames.

September 13, 2018

TbD-net solves the visual reasoning problem by breaking it down to a chain of subtasks. The answer to each subtask is shown in heat maps highlighting the objects of interest, allowing analysts to see the network's thought process.

Artificial intelligence system uses transparent, human-like reasoning to solve problems

Model from MIT Lincoln Laboratory Intelligence and Decision Technologies Group sets a new standard for understanding how a neural network makes decisions.

September 11, 2018

MIT researchers have developed an algorithm that traces intersections in sketches without error. This could save digital artists significant time and frustration when vectorizing an image for animation, marketing logos, and other applications.

Smoothing out sketches’ rough edges

MIT-developed tool improves automated image vectorization, saving digital artists time and effort.

September 11, 2018

2018 MIT Materials Research Laboratory Summer Scholars, are: (front l-r) Michael Molinski, Fernando Nieves Muñoz, Danielle Beatty, Sabrina Shen, Julianna La Lane, Simon Egner; (back l-r) Elizabeth Hallett, Alvin Chang, Sarai Patterson, Ekaterina Tsotsos, Ryan Tollefsen, and Abigail Nason.

Interns at the forefront of new technology

Materials Research Laboratory summer interns tackle materials science challenges, contribute to faculty research labs, and gain new skills.

September 10, 2018

PhD student Lucas Manuelli worked with lead author Pete Florence to develop a system that uses advanced computer vision to enable a Kuka robot to pick up virtually any object.

Robots can now pick up any object after inspecting it

Breakthrough CSAIL system suggests robots could one day be able to see well enough to be useful in people’s homes and offices.

September 10, 2018

Open Access Task Force members consulted with domain experts across campus and beyond to develop an understanding of current local, national, and global practices, policies, and possibilities.

MIT Open Access Task Force releases white paper

Paper provides an overview of efforts to make research and scholarship more freely and openly available.

September 7, 2018

J-PAL North America will partner with the Sacramento-based California Franchise Tax Board to evaluate the impact of strategies to encourage households to file for the California Earned Income Tax Credit (CalEITC).

J-PAL North America announces new partnerships with three state and local governments

New partners will work with J-PAL and its academic researcher network to develop randomized evaluations of programs and policies.

September 6, 2018

New MIT research shows that, unless steady, continuous carbon-free sources of electricity are included in the mix, costs of decarbonizing the electrical system could be prohibitive and end up derailing attempts to mitigate the most severe effects of global climate change.

Study: Adding power choices reduces cost and risk of carbon-free electricity

To curb greenhouse gas emissions, nations, states, and cities should aim for a mix of fuel-saving, flexible, and highly reliable sources.

September 6, 2018

Members of a team developing Adaptable Interpretable Machine Learning at Lincoln Laboratory are: (l-r) Melva James, Stephanie Carnell, Jonathan Su, and Neela Kaushik.

Taking machine thinking out of the black box

Adaptable Interpretable Machine Learning project is redesigning machine learning models so humans can understand what computers are thinking.

September 5, 2018

James Maxwell was the first to realize that light is able to travel in perfect circles within the fish-eye lens because the density of the lens changes, with material being thickest at the middle and gradually thinning out toward the edges.

Fish-eye lens may entangle pairs of atoms

Scientists find a theoretical optical device may have uses in quantum computing.