Credits: Image: Jose-Luis Olivares, MIT
Previous image
Next image
MIT researchers have developed a type of neural network that learns on the job, not just during its training phase. These flexible algorithms, dubbed “liquid” networks, change their underlying equations to continuously adapt to new data inputs. The advance could aid decision making based on data streams that change over time, including those involved in medical diagnosis and autonomous driving.
“This is a way forward for the future of robot control, natural language processing, video processing any form of time series data processing,” says Ramin Hasani, the study’s lead author. “The potential is really significant.”
MIT researchers develop a new ‘liquid’ neural network that’s better at adapting to new info
A new type of neural network that’s capable of adapting its underlying behavior after the initial training phase could be the key to big improvements in situations where conditions can change quickly like autonomous driving, controlling robots or diagnosing medical conditions. These so-called “liquid” neural networks were devised by MIT Computer Science and Artificial Intelligence Lab’s Ramin Hasani and his team at CSAIL, and they have the potential to greatly expand the flexibility of AI technology after the training phase, when they’re engaged in the actual practical inference work done in the field.
It s hoped this new approach could revolutionise technology that relies on decision-making protocols where the data changes over time, or in unpredictable environments, such as medical diagnosis or autonomous driving.
The research will be presented at the AAAI Conference, an artificial intelligence event taking place in Vancouver, Canada, in February. This is a way forward for the future of robot control, natural language processing, video processing - any form of time series data processing, says Ramin Hasani, the study s lead author. The potential is really significant.
Most neural networks have fixed behaviour and they typically don t adjust all that well to changes in incoming data streams. For example, the crash of an Uber autonomous vehicle in 2018 that resulted in the death of Elaine Herzberg, considered the first fatality involving the technology, was said to have been caused by the system being unable to identify the shape of a pedestrian when they were walking alongs
“So, time series data actually create our reality.”
Video processing, financial data, and medical diagnostic applications as everyday examples of time series already central to today’s society.
And although such constantly altered data streams can be unpredictable, analysing these data in real-time, and using them to anticipate future behaviour boasts the potential for further improving technology used every day.
Artificial Intelligence news: Massachusetts Institute of Technology (MIT) engineers have arrived at a new solution to advance AI (Image: Getty)
The MIT team built a neural network engineered to adapt to the variability of real-world systems.
Neural networks are complex algorithms used to recognise patterns after studying sets of “training” examples.