They attempted to technique the condition with different symbolic approaches, in addition to what had been then termed "neural networks"; these were being mostly perceptrons along with other designs which were later on located for being reinventions from the generalised linear models of statistics.[20] Probabilistic reasoning was also used, particu