aboutsummaryrefslogtreecommitdiff
path: root/src/gensvm_base.c
AgeCommit message (Collapse)AuthorFilesLines
2020-07-12Print training accuracyHEADmasterGertjan van den Burg1-0/+3
2017-12-13initialize model size parameters to zeroGertjan van den Burg1-0/+4
2017-02-23Allow setting of the random seed in the modelGertjan van den Burg1-0/+1
2017-02-21Explicitly specify void argument to functions which takeGertjan van den Burg1-2/+2
no arguments.
2017-02-21Minor fixesGertjan van den Burg1-0/+1
2017-02-21Keep a status of trainingGertjan van den Burg1-0/+1
2017-02-21Keep track of elapsed iterations in trainingGertjan van den Burg1-0/+1
2017-02-17allow max_iter to be set in the modelGertjan van den Burg1-0/+1
2017-02-17Remove kernelparam array in favour of explicit kernel parametersGertjan van den Burg1-4/+6
This simplifies a lot of the code and will make it easier to link to other languages.
2016-12-05further unit tests for kernel moduleGertjan van den Burg1-0/+1
2016-11-07compute ZAZ in blocks to increase numerical precisionGertjan van den Burg1-0/+3
2016-10-24update copyright informationGertjan van den Burg1-3/+21
2016-10-14add sparse matrices to GenSVM and reorganize update functionalityGertjan van den Burg1-0/+4
2016-10-06Coding style - set all pointers to NULL initially and after freeGertjan van den Burg1-0/+3
2016-10-06Switch to using dsyrk instead of dsyr for speed.Gertjan van den Burg1-0/+83
Also added a workspace (GenWork) structure for to hold working matrices for the gensvm_get_update() and gensvm_get_loss() functions
2016-09-30Rewrite UU matrix to be K*K*(K-1) instead of n*K*(K-1)Gertjan van den Burg1-4/+1
significant memory reduction by turning the 3D UU matrix into a 2D block matrix, with significantly less dimensions
2016-09-30Remove category matrix from implementationGertjan van den Burg1-6/+0
It's a remnant from the Matlab implementation, where it actually makes sense to use. Here it becomes a significant memory burden for large datasets, and doesn't provide any speedup
2016-09-30remove unnecessary W and t matricesGertjan van den Burg1-9/+0
2016-09-20Fix memory leakGertjan van den Burg1-0/+1
2016-05-16create a single training function for easy external accessGertjan van den Burg1-0/+3
2016-05-16major refactor of the codeGertjan van den Burg1-0/+211