COG SCI 1 Chapter Notes - Chapter 6.3: Machine Learning: Biological Neuron Model, Memory Management, Connectionism
Document Summary
Unlike sam, humans are able to learn something new from what they experience. Schank & childers, 1984: to achieve this level of intelligence, computers need . The same general structures that allow people to make connections. To thoroughly integrate a story into its knowledge structures. To be able to distinguish what is important from what is trivial. To remember things it read in the past just when they are most applicable. Making connections is crucial to pro ting from our experiences. 1950s: ai researchers turned to neural networks, systems making use of connections to produce in the machine something like the dynamic memory within human learning. Earlier ai approaches rely on symbolic processing, solving problems by applying sequences of formal rules for manipulating symbols. High-level programming languages, such as lisp and prolog, facilitate. This approach solves problems at the level of the structure of the machine. Caudill & butler, 1990: there"s a di erence in how the two view info.