They tried to solution the issue with different symbolic strategies, and also what have been then termed "neural networks"; these were being generally perceptrons as well as other types that were afterwards located to get reinventions of the generalised linear designs of stats.[20] Probabilistic reasoning was also employed, especially in automated