graphkit-learn

Commit Graph

Author	SHA1	Message	Date
jajupmochi	80e64a44ae	update pre-image	5 years ago
jajupmochi	b7112cd60f	update pre-image.	5 years ago
jajupmochi	e3985c5481	fitDistance.py: print and save results, update codes of quadratic program. This is the last version with edit costs having no other constraints than being no smaller than 0.	5 years ago
jajupmochi	ee6c79603d	1. update fitDistance.py. 1.1 add parallel computation of GEDs. 1.2 randomly initialize edit costs instead of uniform initialization. 1.3 remove the constrain that sum of the edit costs has to be equal to 1, avoiding sparsity.	5 years ago
jajupmochi	2876f7520a	Update fitDistance.py; constrain edit costs (c_i >= 0, sum(c_i) = 1).	5 years ago
jajupmochi	dd810b92da	Update pre-image.	5 years ago
jajupmochi	8c03f1b049	Update the GED method, use new gedlibpy library.	5 years ago
jajupmochi	edc3db457a	1. allow pre-image algorithm based on IAM to use all k nearest graphs as the initials of IAM. 2. fix a bug: in preimage_iam_random_mix function, sort D_k after updating it.	5 years ago
jajupmochi	57e13c9c5d	Update pre-image.	5 years ago
jajupmochi	344a6f8d4b	Modify function comments of graph kernels.	6 years ago
jajupmochi	b7eef40edf	1. add parallelization choice into the treelet kernel, including "imap_unordered" and None. 2. modify the treelet kernel, use tuples to store canonkeys instead of strings in case that some labels strings contain more than 1 character.	6 years ago
jajupmochi	73b2b7d279	1. add wl subtree kernel. 2. update datasets.	6 years ago
jajupmochi	92c74448ec	add the treelet kernel.	6 years ago
jajupmochi	a13dd94f6a	test of a new preimage algorithm.	6 years ago
jajupmochi	7b75c1fc97	1. add verbose switch. 2. fix two dimension bugs in model selection function when computing std.	6 years ago
jajupmochi	ddd921696a	rearrange notebooks directory, correct all scripts.	6 years ago
jajupmochi	6b4738df30	rearrange pygraph/kernels file.	6 years ago

17 Commits (0cc89ae026108889b986aba01e3f999d12d37a2b)