id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_dksywpuce5hyvazldbmanbc62i	Facundo Bromberg	Efficient Markov Network Structure Discovery Using Independence Tests	2009	36	.pdf	application/pdf	18598	2745	82	Both algorithms use statistical independence tests to infer the structure by successively constraining the set of structures consistent with the results of these Until very recently, algorithms for structure learning were based on maximum likelihood estimation, which has been proved to be NP-hard for Markov networks due to the We present two algorithms for MN structure learning from data: GSMN∗ (Grow-Shrink Markov Network learning algorithm) and GSIMN (Grow-Shrink Inference-based Markov and GSIMN algorithms presented apply to any case where an arbitrary faithful distribution can be assumed and a probabilistic conditional independence test for that distribution (Grow-Shrink Inference-based Markov Network learning algorithm), introduced in this section, uses the Triangle theorem in a similar fashion to extend GSMN∗ by inferring the value For each data set and each algorithm, we report the weighted number of conditional independence tests conducted to discover the network and the accuracy, as defined	./cache/work_dksywpuce5hyvazldbmanbc62i.pdf	./txt/work_dksywpuce5hyvazldbmanbc62i.txt