Cognitive behavioral institute researchers have found that a form of plasticity they call “deep learning” can improve our cognitive abilities and make us better at learning new things.
Key points:The new form is called “deep-learning” and it involves training your brain to use a set of specific cognitive patternsDeep learning is a form in which the brain learns to apply specific kinds of mental skills to new tasksA study published in the journal Nature Neuroscience found that people who received training in the form of deep-learning were more successful in their ability to learn new tasks.
Researchers from the Cognitive Neuroscience Institute at the University of Cambridge in the UK conducted a meta-analysis of studies investigating the effects of training in a deep-training paradigm.
Researchers found that deep-trained participants outperformed controls on a variety of tasks, including learning to identify and respond to faces, and visual recognition.
“Deep learning can be thought of as the brain’s own version of the visual processing task, and in the last decade, it has been widely recognised that the brain uses different types of neural processing to process visual information,” said lead author, Professor Andrew Tully, from the Centre for the Study of Consciousness.
“This research suggests that deep learning is particularly important in the areas of cognitive control and memory, which are critical for learning, as they are critical to understanding our thoughts and behaviour.”
Dr Tully and his colleagues tested the effect of training participants in the deep- learning mode on a wide variety of different tasks.
The team looked at studies that compared people with and without deep-learned abilities.
They then applied the technique to two types of tasks: the familiar recognition task (FRC), which measures whether you can recognize the face of a person you have previously met, and the non-recognition task (NRT), which assesses whether you know the name of a song, or the number of words in the phrase “A few things in my life are about to change”.
Deep learning was able to improve the recognition task in both groups.
The researchers also found that participants with deep- learned abilities had significantly better performance on NRT, and more successful performance on the familiar face recognition task.
These findings suggest that deep training can improve performance in a wide range of tasks.
Dr Tullie said the findings of this study suggest that the benefits of deep learning could also apply to learning new tasks, as the researchers have previously shown.
“It’s a good example of the way in which deep learning has the potential to improve cognition, even when it’s not being used in an everyday context,” he said.
“We also found this to be true for tasks that were previously thought to be very difficult to learn.”
Dr Trigg, from Cambridge’s Cognitive Neuroscience Centre, said it was exciting to see the benefits from training.
“One of the main things we’re trying to do is to apply what we have learned in neuroscience to the everyday world,” he explained.
“When you’re trying out this new form, we’ve found that it’s very useful for people in real-world situations, such as driving or building a house.”
What we want to do now is to show that deep neural learning is also relevant in the real world, and that the training is beneficial for people who are not using it in the everyday context.
“Ultimately, we hope that our results will lead to improved technology for both human and robotic learning.”