Toward speech recognition for uncommon spoken languages
Reducing the complexity of a powerful machine-learning model may help level the playing field for automatic speech-recognition around the world.
Avoiding shortcut solutions in artificial intelligence
A new method forces a machine learning model to focus on more data when learning a task, which leads to more reliable predictions.
Taming the data deluge
A National Science Foundation-funded team will use artificial intelligence to speed up discoveries in physics, astronomy, and neuroscience.
3 Questions: Investigating a long-standing neutrino mystery
Graduate student Nicholas Kamp describes the MicroBooNE experiment and its implications for our understanding of fundamental particles.
Making machine learning more useful to high-stakes decision makers
A visual analytics tool helps child welfare specialists understand machine learning predictions that can assist them in screening cases.
Saving seaweed with machine learning
PhD candidate Charlene Xia is developing a low-cost system to monitor the microbiome of seaweed farms and identify diseases before they spread.
At Mass STEM Week kickoff, MIT RAISE announces Day of AI
Artificial intelligence is top-of-mind as Governor Baker, President Reif encourage students to “see yourself in STEM.”
Artificial networks learn to smell like the brain
When asked to classify odors, artificial neural networks adopt a structure that closely resembles that of the brain’s olfactory circuitry.
Accelerating the discovery of new materials for 3D printing
A new machine-learning system costs less, generates less waste, and can be more innovative than manual discovery methods.
These neural networks know what they’re doing
A certain type of artificial intelligence agent can learn the cause-and-effect basis of a navigation task during training.
A dispatch and routing platform to improve deliveries
Wise Systems has grown from an MIT class project to a company helping multinationals improve last-mile logistics.
Deep learning helps predict traffic crashes before they happen
A deep model was trained on historical crash data, road maps, satellite imagery, and GPS to enable high-resolution crash maps that could lead to safer roads.
3 Questions: Kalyan Veeramachaneni on hurdles preventing fully automated machine learning
Researchers hope more user-friendly machine-learning systems will enable nonexperts to analyze big data — but can such systems ever be completely autonomous?
Blockchain technology could provide secure communications for robot teams
The transaction-based communications system ensures robot teams achieve their goal even if some robots are hacked.