yogsototh/clj-ml
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
218 commits 1 branch 0 tags 406 KiB
Commit graph

138 commits

Author SHA1 Message Date
Joshua Eckroth
cce83d924d Bugfix. 2013-07-04 09:36:31 -04:00
Joshua Eckroth
623e9a1ef3 New format for terms on public datasets. Bumped to version 0.3.12. 2013-07-04 09:34:53 -04:00
Joshua Eckroth
4a4ef5ea03 Starting support for public datasets. Added reuters21578 dataset handling. 
Bumped to version 0.3.11.
 2013-07-04 09:00:48 -04:00
Joshua Eckroth
6f03716d0a Added function clj-ml.data/docs-to-dataset to support translating text documents (with title, fulltext, and terms) into wordvec datasets for binary classification. 
Bumped to version 0.3.10.
 2013-07-04 08:53:45 -04:00
Joshua Eckroth
edceff891b Fixed bug reading instance attribute values. 
Bumped to 0.3.8.
 2013-07-01 14:08:01 -04:00
Joshua Eckroth
a8ccca6d53 Support stemmers (use Porter/English/etc. rather than lowercase form). 
Bumped to version 0.3.7.
 2013-07-01 14:07:34 -04:00
Joshua Eckroth
2ce2d41885 Capture log messages from libsvm (which is noisy). 2013-06-21 01:20:40 -04:00
Joshua Eckroth
4aa08595a1 More fixes for svm grid operation; added libsvm to project deps. 2013-06-21 01:00:55 -04:00
Joshua Eckroth
52a1de6732 Bugfixes for new svm grid classifier. 2013-06-21 00:32:46 -04:00
Joshua Eckroth
f9704e67a4 Added libsvm-grid classifier type, which is really a collection of classifiers. 2013-06-21 00:23:24 -04:00
Joshua Eckroth
709b551bfd Support LibSVM classifier. Requires libSVM.jar to be in classpath. 2013-03-30 18:16:47 -04:00
Joshua Eckroth
535c4b1222 Support loading/saving libsvm data. 2013-03-30 18:16:33 -04:00
Joshua Eckroth
aab3d42a99 Don't print evaluation results, just collect them. 2013-03-25 16:27:33 -04:00
Joshua Eckroth
05cf721ec5 Changed resample filter to resample-unsupervised, and added resample-supervised. 2013-03-22 12:35:24 -04:00
Joshua Eckroth
54eb9cf78b Added resample filter. 2013-03-22 10:58:03 -04:00
Joshua Eckroth
7004dd87cd Added :reorder-attributes filter. 2013-03-22 10:09:38 -04:00
Joshua Eckroth
ed3bf1d71d Added StringToWordVector filter. 2013-03-22 07:55:01 -04:00
Ben Mabey
aacc60b8b1 adds pki-unsupervised-discretize filter wrapper 2013-02-05 20:09:29 -07:00
Ben Mabey
4f13798e80 blow up when the attribute can't be found in attribute-value-fn 2012-03-27 15:17:56 -06:00
Ben Mabey
dc12081b4a attaches instance weight info as metadata when converting to a map 2012-03-26 12:20:45 -06:00
Ben Mabey
042de1895a updates project to clojure 1.3, removes contrib and autodoc deps 2012-01-30 15:27:54 -07:00
Ben Mabey
239ef8a6f8 removes stray require 2012-01-23 16:51:15 -07:00
Ben Mabey
096b9f9b03 adds attribute-value-fn helper fn 2012-01-23 16:38:36 -07:00
Ben Mabey
e5398d7c87 fixes bug in add-attribute filter 2012-01-23 16:38:33 -07:00
Ben Mabey
cd79db2b90 more simple cleanups to the option conversion code 2012-01-05 15:11:00 -07:00
Ben Mabey
199d58209e more loop -> reduce conversion... 
someone should tell the original author about reduce.. it would save him
a lot of time. :)
 2012-01-05 15:10:59 -07:00
Ben Mabey
e0e80f42c3 don't add an option flag if the value is false 2012-01-05 14:39:00 -07:00
Ben Mabey
c25d39521b removes mongodb store.. I never used it and won't be supporting it 2012-01-05 13:29:45 -07:00
Ben Mabey
e9b63b877f adds map-fast-vec helper 2011-12-13 20:10:44 -07:00
Ben Mabey
fa2d39ffe5 change the dataset's name when randomized (like the filters do) 2011-12-13 20:10:11 -07:00
Ben Mabey
8a01361512 adds take-dataset 2011-12-08 15:28:27 -07:00
Ben Mabey
d313f621d1 adds remove-range filter so split-dataset can take a number to split on 2011-12-08 15:13:34 -07:00
Ben Mabey
b454699af2 adds copy-dataset and randomize-dataset 2011-12-08 15:13:10 -07:00
Ben Mabey
6f8205d936 adds randomize-dataset! 2011-12-07 11:23:43 -07:00
Ben Mabey
0c32c318af adds split-dataset via new remove-percentage filter wrapper 2011-12-07 11:15:19 -07:00
Ben Mabey
84b1903b63 adds simple wrappers for Weka's attribute (feature) selection package 2011-10-04 13:10:01 -06:00
Ben Mabey
6f5c100d95 extracts Weka Options helpers into own ns due to circ dep issue 2011-08-29 16:50:40 -06:00
Ben Mabey
fb046172e5 Merge remote branch 'lt/master' 2011-07-14 16:12:38 -06:00
Ben Mabey
539bf83440 uses weka's built in enumerateInstances for dataset-seq 2011-07-14 16:12:13 -06:00
Ronny Bjarnason
0641b48809 dataset-weights returns weights of dataset instances 2011-07-14 14:28:46 -06:00
Ronny Bjarnason
2cb2ccd574 add classifier FastRandomForests 2011-04-22 15:53:39 -06:00
Ben Mabey
953cc0bac5 adds needed type hints in the filter factory fn 2011-04-18 14:21:16 -06:00
Ben Mabey
7357a720e9 makes representation of NaNs (no values) symmetric with creation- nil 2011-04-05 13:55:21 -06:00
Ben Mabey
a83e21d72b allows for creation of instances (and datasets) with nil values 2011-04-05 13:34:23 -06:00
Ben Mabey
94ffacaf55 removes cavepaint 2011-04-05 13:31:21 -06:00
Ben Mabey
eb3937eb6f adds needed doto call on dataset-set-name so ds is returned 2011-04-01 14:37:48 -06:00
Ben Mabey
fb02b300c3 Merge branch 'master' of github.com:leadtune/clj-ml 2011-03-24 11:46:18 -06:00
Ben Mabey
ea65e5d973 allows save-instances to take an OutputStream 2011-03-24 11:35:20 -06:00
Ronny Bjarnason
b4ff80cc98 SPegasos: set loss function in options 2011-03-21 15:31:26 -06:00
Ronny Bjarnason
2709ececdb fix typo 2011-03-17 14:01:30 -06:00