Philosophers have long emphasized the active nature of perception and the intimate relation between action and cognition. Essentially, cognitive behavior results from our interaction with our environments.
Our environments are filled with a plethora of stimuli. To "discover" them, we have to constantly discriminate between relevant and irrelevant features. Many of our actions are based off of sensory reactions to inputs, which reflect a specific task, i.e., sensorimotor relations (laws) that are learned and reinforced via experience.
Scientists have studied long-term changes of sensorimotor neural representations obtained during habit learning. The part of the brain responsible for this learning is the corpus striatum (part of the basal ganglia of the brain - caudate and lentiform nuclei), which receives direct cortical input.
Unsupervised learning happens in the cortex while reinforcement learning occurs in the basal ganglia. The cortex pre-processes data to yield a representation that is suitable for reinforcement learning by the basal ganglia.
The seven deep brain nuclei of the basal ganglia are involved in a variety of crucial brain functions and are tightly linked to the dopaminergic neuromodulatory system, which plays a fundamental role in predicting future rewards and punishment.
Essentially, reward-based learning models are guided by trial and error, as the brain continually maps (makes connections) between states and actions that yield the maximal future reward.
Reward-based learning is most complimented the Nomadic learning environment, where information changes based on geological location. It's akin to learning how to "think on our feet."
Our brains easily learn goal-relevant features within a single unified framework, but when these models fail to discriminate between action-relevant and irrelevant features, they hinder the brain's learning architecture to the 2-dimensional plane. Once the sensorimotor laws have been learned and the visual features for navigation have been captured, we are able to navigate.
Notwithstanding, an educational model that incorporates new geological locations offers a clear advantage to fine tuning our ability to learn and adapt. Variety is a fundamental experience in exploring different sensory channels. Stagnant learning environments that are geologically limited have too narrow a field-of-view and do not make use of the brain's learning circuitry, which easily adapts to new inputs, resulting in increased cognitive functioning.
Our brains easily learn goal-relevant features within a single unified framework, but when these models fail to discriminate between action-relevant and irrelevant features, they hinder the brain's learning architecture to the 2-dimensional plane. Once the sensorimotor laws have been learned and the visual features for navigation have been captured, we are able to navigate.
Notwithstanding, an educational model that incorporates new geological locations offers a clear advantage to fine tuning our ability to learn and adapt. Variety is a fundamental experience in exploring different sensory channels. Stagnant learning environments that are geologically limited have too narrow a field-of-view and do not make use of the brain's learning circuitry, which easily adapts to new inputs, resulting in increased cognitive functioning.
No comments:
Post a Comment