tekhnofiend
|
11cce6677a
|
added pica
|
2008-04-06 01:11:52 +00:00 |
|
houyang
|
f8930f2373
|
rvm classification
|
2008-04-04 19:57:31 +00:00 |
|
houyang
|
dee61ae4f7
|
Relevance Vector Machine (RVM)
|
2008-03-31 20:26:46 +00:00 |
|
houyang
|
a32ffc6cbc
|
Add a statistics module. More functions to contribute.
|
2008-03-31 20:22:28 +00:00 |
|
houyang
|
bf8a7b51b7
|
SVM regression
|
2008-03-31 20:14:42 +00:00 |
|
houyang
|
90fdbe203b
|
Add a general cross validation class, which can be used for classification, regression and density estimation. Alpha Version. Still under construction.
|
2008-03-31 20:07:15 +00:00 |
|
tekhnofiend
|
6c3e22af6d
|
|
2008-03-30 16:51:24 +00:00 |
|
tekhnofiend
|
7f6cdcd037
|
|
2008-03-30 16:50:03 +00:00 |
|
tekhnofiend
|
e122850378
|
updated
|
2008-03-30 16:24:35 +00:00 |
|
houyang
|
e0436b7ffd
|
|
2008-03-21 23:23:32 +00:00 |
|
gmravi2003
|
7a7445d2d0
|
Fully functional. No leaks
|
2008-03-09 22:13:26 +00:00 |
|
gmravi2003
|
46f9df9c89
|
Slightly modified. Still workingb but with mem leaks
|
2008-03-09 00:05:52 +00:00 |
|
gmravi2003
|
cd564b8a9f
|
Fully working. However there are mem leaks
|
2008-03-08 23:05:13 +00:00 |
|
Dongryeol Lee
|
dc9cea5bd7
|
Another fix to the Compute function
|
2008-03-03 20:55:54 +00:00 |
|
Dongryeol Lee
|
93fdb3c60b
|
The naive code had a bug in computing the influence value of a point; it should have computed t(q)^T (B^T W(q) B)^{-1} t(q) W(0), instead of t(q)^T (B^T W(q) B)^{-1} t(q)
|
2008-03-03 03:48:36 +00:00 |
|
gmravi2003
|
dc356ba601
|
Complete. Will need to do a small final check
|
2008-03-03 00:51:51 +00:00 |
|
gmravi2003
|
026a85eacc
|
partially complete
|
2008-03-02 16:18:44 +00:00 |
|
gmravi2003
|
289f65b6d1
|
Partially complete code
|
2008-03-02 07:19:22 +00:00 |
|
Dongryeol Lee
|
bb0441fb51
|
Fix in the compute function for naive code which contained erroneous extra argument
|
2008-03-02 06:12:24 +00:00 |
|
Dongryeol Lee
|
c7122cb14c
|
Caught a serious bug in non-shuffling of the weight diagram magnitude and influence values in the dense version of the algorithm
|
2008-03-02 04:42:49 +00:00 |
|
gmravi2003
|
4bf3d47d76
|
Looks to be fine
|
2008-02-28 02:18:10 +00:00 |
|
tekhnofiend
|
38987f4c15
|
updated niche/functional
|
2008-02-27 07:04:19 +00:00 |
|
tekhnofiend
|
e917e96281
|
updated niche/functional
|
2008-02-27 07:03:39 +00:00 |
|
tekhnofiend
|
442be9e768
|
updated niche/functional
|
2008-02-26 19:43:52 +00:00 |
|
tekhnofiend
|
996d016cc3
|
updated niche/functional
|
2008-02-26 19:42:49 +00:00 |
|
gmravi2003
|
533d3a5c4f
|
Looks like working.Havent tested it completely
|
2008-02-25 16:24:08 +00:00 |
|
tekhnofiend
|
f33bc0144a
|
added k_fold_fpc_cv* to niche/functional
|
2008-02-25 03:12:12 +00:00 |
|
tekhnofiend
|
449c2cc18b
|
updated niche/functional
|
2008-02-25 03:10:43 +00:00 |
|
Dongryeol Lee
|
2b5b4a2e0e
|
Fix to the variable bandwidth case
|
2008-02-24 20:02:00 +00:00 |
|
tekhnofiend
|
4e35eb39f0
|
added results_gene to niche/functional
|
2008-02-24 07:10:44 +00:00 |
|
tekhnofiend
|
9bdc54c492
|
updated niche/functional
|
2008-02-24 07:09:00 +00:00 |
|
Dongryeol Lee
|
f384336b14
|
Added the multiindex expansion tool
|
2008-02-24 06:17:29 +00:00 |
|
Dongryeol Lee
|
0f0c837f7d
|
Added the naive version
|
2008-02-24 06:16:34 +00:00 |
|
Dongryeol Lee
|
1eb548f9ac
|
Initial check in of dense version
|
2008-02-24 06:16:13 +00:00 |
|
tekhnofiend
|
71f2f6c960
|
updated niche/functional
|
2008-02-24 02:06:01 +00:00 |
|
gmravi2003
|
43bc1068c7
|
Fully working without adaptive BW
|
2008-02-23 00:43:40 +00:00 |
|
Parikshit Ram
|
710db82a9e
|
optimizers tried with mu > 1 but exploded so reduced it so i dont know exactly whats the change in optimizers.h, but the main.cc is added to the svn
|
2008-02-22 19:40:28 +00:00 |
|
Parikshit Ram
|
c54699dd46
|
optimizers added with more terminating conditions although none seem to work, and main deleted accidentally
|
2008-02-22 06:47:01 +00:00 |
|
Parikshit Ram
|
c684ed2a5a
|
different terminating conditions included
|
2008-02-22 00:11:18 +00:00 |
|
Parikshit Ram
|
529e4a51eb
|
SGD and SMD_single step implemented but some problem with SMDSS
|
2008-02-20 22:23:26 +00:00 |
|
gmravi2003
|
0671be4b6d
|
KNN Based regression. Confidence Intervals are still being added
|
2008-02-20 20:23:06 +00:00 |
|
Parikshit Ram
|
08c67b37a8
|
gradient descent added
|
2008-02-20 09:12:36 +00:00 |
|
Parikshit Ram
|
9dfad1facd
|
faster folders removed
|
2008-02-20 04:25:31 +00:00 |
|
Parikshit Ram
|
1d792308a3
|
corrected version updated
|
2008-02-20 04:11:48 +00:00 |
|
Parikshit Ram
|
f6c20f0fe5
|
corrected stuff working
|
2008-02-20 04:08:09 +00:00 |
|
tekhnofiend
|
a68ae84ad3
|
updated niche/functional
|
2008-02-20 01:51:38 +00:00 |
|
tekhnofiend
|
6e2a04cb3d
|
updated niche/functional
|
2008-02-19 16:55:40 +00:00 |
|
tekhnofiend
|
c8a16a96f6
|
updated niche/functional
|
2008-02-19 16:51:40 +00:00 |
|
tekhnofiend
|
b8ca6ba755
|
updated niche/functional
|
2008-02-19 05:49:53 +00:00 |
|
tekhnofiend
|
dc10577b75
|
updated niche/functional
|
2008-02-19 01:01:13 +00:00 |
|