Familiar Problem Space (FPS): A Novel Approach for Brain-like Learning

Abstract- Lazy learning methods search for the match, while there may be no exact match, so the best match is called for. Unfortunately, there is no general way to search memory for the best match without examining every element of memory. It is proven that brain can not traverse more than 100 neurons in less then 200 milliseconds which we need to solve most of our routine decisions. This paper (inspired by theories of cognition and human brain learning) explains the model of brain learning and then propose a new strategy to learn and classify examples with additional advantages over traditional Back-propagation Neural Network