id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_k36ubeddt5fuvoo5aum6ofmeuq	Stan C. Kwasny	Rule-based training of neural networks	1991	12	.pdf	application/pdf	7882	905	66	explore two types of learning, deductive and inductive, in the context of a rule-based, deterministic Typically, these systems develop from consultations with recognized experts who provide knowledge and experience during the process of rule development and debugging. (see Fanty, 1985; Selman & Hirst, 1985; Waltz & Pollack, 1985). set of deterministic grammar rules and tested with sentences which are grammatical and ones that are not. the rules or sentences on which its training is based Experiments are conducted to determine the effectiveness of training and to investigate whether the connectionist network generalizes properly to ungrammatical and lexically ambiguous cases. As mentioned earlier, there are two distinct approaches to training a network to parse sentences. Rule-Based Training of Neural Networks 57 Rule-Based Training of Neural Networks 57 Rule-Based Training of Neural Networks 57 Rule-Based Training of Neural Networks 57 Rule-Based Training of Neural Networks 57 Rule-Based Training of Neural Networks 57	./cache/work_k36ubeddt5fuvoo5aum6ofmeuq.pdf	./txt/work_k36ubeddt5fuvoo5aum6ofmeuq.txt