A neuroscientist from New York has offered an explanation as to how the human brain formulates perceptual prediction.
Prediction or forecasting is an essential function of the brain. Research has shown that a lack of forecasting would result into delayed neural processing.
Forming Predictions
In a new study, Professor David Heeger of New York University proposes a new outline which would very well explain the way human brains make predictions.
"It has long been recognized that the brain performs a kind of inference, combining sensory information with expectations," said Heeger.
The neuroscientist explained that the brain's neural networks represent a kind of model of our surroundings. Theories relating to the function of the human brain and neural networks, especially ones that artificial intelligence employ, are hierarchal in structure. As sensory input comes in at one end, gradually more immaterial representations would be calculated along the process.
Heeger explains that taking an abstract representation on top of the hierarchy, and running it backwards from top to bottom through the neural net would generate something similar to a sensory expectation.
The neural network in the brain processes in a manner which is feedforward: sensual input to abstract depiction similar to archetypal artificial intelligence neural networks.
Additionally, A.I. cannot completely capture these phenomena despite the fact it seeks to duplicate human decision making and processing, it's because it's solely ignoring prediction, exploration and plus it draws from a feedforward structure.
Heeger said how the brain operates to make predictions is obscure to us yet. We've also misunderstood the models similar to meteorologists' employment of brain function. Forecasters make their predictions by depending on past weather information, as to plan climate conditions with the following days.
Sensory Forecast
To generate a such a manner, Heeger's theory would instead run in a feedback mode, or it would run in a mode where there's both memory recall and mental imagery, whereby inferences blend sensory with forecast. Overall, the theory simply mixes inference with prediction.
Heeger speculates that the exploration, may it be noise-driven or not, may be the neural footing of ingenuity. The total quantity of exploration would be determined by the noise amount, and is controlled by neuromodulators and oscillations in the brain activity.
These factors would explain why there are no particular genetic mechanisms that cause autism, but instead, individuals with this neurocomputational deficit have their brains functioning improperly. He added that scientists hope to understand and shed light about these neurocomputational deficits someday.
Details of the study are published in the journal PNAS.
