On the computational basis of learning and cognition: Arguments from LSA

摘要: Publisher Summary This chapter discusses the computational basis of learning and cognition. To deal with a continuously changing environment, living things have three choices: (1) evolve unvarying processes that usually succeed, (2) genetically fixed effector, perceptual, functions are contingent on (3) learn adaptive during their lifetimes. The theme this is relation between (3): nature evolutionarily determined support learning. principal goal has been to suggest high-dimensional vector space computations based empirical associations among very large numbers components could be close model fundamental most in both verbal perceptual domains. More powerful representational effects can brought about by linear inductive combinations elements vocabularies than often realized. Success one such demonstrate many natural properties language commonly assumed essentially more complex, nonlinear, and/or unlearned, along evidence argument similar may serve roles object recognition, taken reaffirm possibility single underlying associational mechanism lies behind special complex appearing cognitive phenomena.