id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_etgjv3izcvcoxa7feakkgtd764	Stefan L. Frank	Sentence-processing in echo state networks: a qualitative analysis by finite state machine extraction	2010	22	.pdf	application/pdf	11775	2036	80	Sentence-processing in echo state networks: a qualitative analysis by finite analysis by finite state machine extraction', Connection Science,, First published on: 22 March 2010 (iFirst) It has been shown that the ability of echo state networks (ESNs) to generalise in a sentence-processing machines (FSMs) from the data about the inputs, internal states, and outputs of recurrent neural networks that process symbol sequences. extracting finite state machines (FSMs) from network's input symbols, recurrent-layer states, and probabilistic FSM given the network's input, internal states, and outputs. To merge and split the state space, CrySSMEx creates a number of simple vector quantisers, arranged hierarchically in a graph (Figure 8 in Section 4.3 shows an example). network predicts the following input to be a noun, and indeed, all sentences of the language start After receiving the sentence's second word (a verb), the FSM moves to State 2, where it predicts	./cache/work_etgjv3izcvcoxa7feakkgtd764.pdf	./txt/work_etgjv3izcvcoxa7feakkgtd764.txt