fobos_clj

Implementation of Forward Backward Splitting in clojure.

Usage

Following is how to use the document classifier:

$ wget http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/binary/news20.binary.bz2
$ bzip2 -d news20.binary.bz2
$ ruby shuffle news20.binary > tmp.txt
$ head -n 15000 tmp.txt > train.txt; tail -n 4996 tmp.txt > test.txt
$ lein run --help
$ lein run --mode train --model fobos_clj.logistic.Logistic --train-filename train.txt --max-iter 100 --eta 1.0 --lambda 0.005 --model-file weight.model
$ lein run --mode test --test-filename test.txt --model-file weight.model

Using from clojure

(let [examples [[1 [[1 1] [2 1]]]
                [-1 [[1 2] [2 0]]]]
      init-weight {0 0.1, 1 0.2, 2 0.5, 3 -0.1}
      iter 10
      eta 0.1
      lambda 0.1
      svm (.update-weight ;; train
           (fobos_clj.svm.SVM. examples init-weight eta lambda) iter)]
  ;; test
  (.classify svm [[1 1] [2 1]]))

License

Copyright (C) 2012 Yasuhisa Yoshida

Distributed under the Eclipse Public License, the same as Clojure.