238 Commits (master)

Author SHA1 Message Date
  jajupmochi ed78e65dd8 [Features] Add model seletion methods with validation set: KFoldWithValid, RepeatedKFoldWithValid, BaseCrossValidatorWithValid. Required version of scikit-learn is upgraded to 1.1.0, to support the input_name argument of sklearn.utils.check_array used in gklearn.model_selection._split.py. 3 years ago
  jajupmochi b36eaae177 [Features] Add model seletion methods with validation set: KFoldWithValid, RepeatedKFoldWithValid, BaseCrossValidatorWithValid. Required version of scikit-learn is upgraded to 1.1.0, to support the input_name argument of sklearn.utils.check_array used in gklearn.model_selection._split.py. 3 years ago
  jajupmochi 5eb90655a4 [Features] Add model seletion methods with validation set: , , . Required version of scikit-learn is upgraded to 1.1.0, to support the argument of used in . 3 years ago
  jajupmochi a76335ed16 [Features][API Changes] Update kernel classes. 3 years ago
  jajupmochi 5e37d4447f [Major Features] Add GEDModel which is compatibale with . 3 years ago
  jajupmochi a7e189134d [Fix] Change the output directory name. 3 years ago
  jajupmochi 1946d46964 [Feature] Allow to permutate nodes in graphs when using bipartite to estimate GED. This feature is implementated in the method in Python, which invokes GEDLIB in C++ by Cython. 3 years ago
  jajupmochi 42acfd0236 [Fix] Fix laplacian_kernel and cosine_kernel between pairs of numpy vectors. 3 years ago
  jajupmochi dd58f602ec [Feature] Add kronecker_delta_kernel function between a pair of numpy vectors. 3 years ago
  jajupmochi 7d3e929b36 [Feature] Add several new kernels between vectors. 3 years ago
  jajupmochi 310b01c347 [Major Feature] WLSubtree kernel can now deal with symbolic edge labels (for Gram matrix computation without parallelization). Meanwhile, the dummy labels are no longer added a priori. 4 years ago
  jajupmochi bd8bf2b234 [Enhancement] Add more verbose for the WL kernel. 4 years ago
  jajupmochi 609c8c1518 [Enhancement] Allow deciding whether or not to make a copy of input graphs in GraphKernel class. 4 years ago
  jajupmochi 785a5c1926 [To finish] Remove dummy labels in treelet kernel (for the contest.) 4 years ago
  jajupmochi 231c050f88 [To finish] Remove dummy labels in treelet kernel (for the contest.) 4 years ago
  jajupmochi 23a938482a [[Enhancement] gklearn.utils.normalize_gram_matrix function now raises… 4 years ago
  jajupmochi 4fd314d374 [Fix] Fixed the bug that gklearn.utils.kernels.gaussiankernel can not handle lists as inputs. 4 years ago
  jajupmochi 1800923c45 [API change] gklearn.utils.kernels.gaussiankernel function no longer transforms inputs to numpy.array of float values. This may be faster, but may cause some errors to the previous codes. 4 years ago
  jajupmochi 70a11dfd1b [Enhancement] gklearn.utils.normalize_gram_matrix function now raises an excepetion if the diagonal includes negative values.\n[Enhancement] gklearn.utils.get_graph_kernel_by_name function now accepts strings of kernel names the same as the keys of gklearn.kernels.GRAPH_KERNELS and supports more graph kernels. 4 years ago
  jajupmochi b843707e9c [Enhancement] gklearn.dataset.Dataset class can now automatically get the task type of the given dataset (regression or classification). 4 years ago
  jajupmochi 325f51cb6f [Feature] Add the gklearn.model_learning.dichotomous_permutation function. 4 years ago
  jajupmochi ed17010535 [Major Feature][Todo] Add model_learning module, including a new NestedCV class and a Workflow class for graph kernel computation. 4 years ago
  jajupmochi bb36d0f507 [Major Feature] Graph kernel classes now can compute kernel matrix between two different list of graphs using fit/transform methods which uses the same scheme as the scikit-learn library! 4 years ago
  jajupmochi d08eddc69c [Enhancement] The global variable now uses kernel classes as values. 4 years ago
  jajupmochi 679953200c [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 81a881bd79 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 14bd9db5a7 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 29bcfe4dab [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 13c0fb37c1 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 6e898240e3 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 27de2454a9 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 25e10c8d49 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi cba8047279 [Exp] Update computation of ged stability. 4 years ago
  jajupmochi 59687d1e87 [Exp] Add taskhub to run exps all in once. 4 years ago
  jajupmochi 74bab60deb [Exp] Add taskhub to run exps all in once. 4 years ago
  jajupmochi 13ac6d9153 [Exp] Add taskhub to run exps all in once. 4 years ago
  jajupmochi 756371aaf1 [Exp] Add taskhub to run exps all in once. 4 years ago
  jajupmochi ac78c8517a [Exp] Add taskhub to run exps all in once. 4 years ago
  jajupmochi 07478a5717 [Exp] Add taskhub to run exps all in once. 4 years ago
  jajupmochi e991f59742 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 71df9f591f [Update] Move the import of control out of the __init__(). 4 years ago
  jajupmochi 47951de049 [Update] Move the import of control into the __init__(). 4 years ago
  jajupmochi 684d26bf7e [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi bfe64a1f66 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 6332e5684d [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 4bfde8d881 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi 59a36aa0a0 [Exp] Update exceptions in fcsp exps. 4 years ago
  jajupmochi a638bfa45a [Exp] FCSP: 1. Use no more than 10 tlong tasks. 2. use ljia instead of explicit name. 3. My first SCI journal paper is accepted! 4 years ago
  jajupmochi 8c6168b8f8 [Exp] Update special tasks. 4 years ago
  jajupmochi e5b977734c [Fix bugs] length in get_iters in compute_geds(). 4 years ago