Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
Upcoming SlideShare
画像処理でのPythonの利用
Next

Share

SHOGUN使ってみました

第1回:はじめてのSHOGUN編

Related Books

Free with a 30 day trial from Scribd

See all

SHOGUN使ってみました

  1. 1. SHOGUN 2011 4 23 9 CV PRML @yasutomo57jp ( @inco_san )
  2. 2. SHOGUN 1 SHOGUN 2011 4 23 9 CV PRML @yasutomo57jp ( @inco_san )
  3. 3. SHOGUN
  4. 4. * OpenCV * http://d.hatena.ne.jp/takmin/20110306/1299423617
  5. 5. • SHOGUN
  6. 6. • SHOGUN • 1 SHOGUN
  7. 7. • SHOGUN • 1 SHOGUN • Static Interface
  8. 8. • SHOGUN • 1 SHOGUN • Static Interface • 2 SHOGUN
  9. 9. • SHOGUN • 1 SHOGUN • Static Interface • 2 SHOGUN • Modular Interface
  10. 10. • SHOGUN • 1 SHOGUN • Static Interface • 2 SHOGUN • Modular Interface • 3 C++ ( )
  11. 11. • SHOGUN • 1 SHOGUN • Static Interface • 2 SHOGUN • Modular Interface • 3 C++ ( ) • libshogun
  12. 12. • SHOGUN • 1 SHOGUN • Static Interface • 2 SHOGUN • Modular Interface • 3 C++ ( ) • libshogun
  13. 13. • SHOGUN • 1 SHOGUN • Static Interface • 2 SHOGUN • Modular Interface • 3 C++ ( ) • libshogun
  14. 14. SHOGUN
  15. 15. SHOGUN •
  16. 16. SHOGUN • • SVM !
  17. 17. SHOGUN • • SVM ! • SVM OCAS, Liblinear, LibSVM, SVMLight, SVMLin, GPDT
  18. 18. SHOGUN • • SVM ! • SVM OCAS, Liblinear, LibSVM, SVMLight, SVMLin, GPDT • Linear, Polynomial, Gaussian and Sigmoid Kernel
  19. 19. SHOGUN • • SVM ! • SVM OCAS, Liblinear, LibSVM, SVMLight, SVMLin, GPDT • Linear, Polynomial, Gaussian and Sigmoid Kernel •
  20. 20. SHOGUN • SVM !! • LDA : Linear Discriminant Analysis • LPM : Linear Programming Machine • (Kernel) Perceptron • HMM
  21. 21. SHOGUN • •
  22. 22. SHOGUN
  23. 23. Q. Matlab
  24. 24. Q. Matlab
  25. 25. Octave
  26. 26. Python
  27. 27. Python
  28. 28. Q. C++ …
  29. 29.
  30. 30.
  31. 31. SHOGUN
  32. 32. SHOGUN
  33. 33. • Static Interface • • • Modular Interface • Python Octave • • libshogun • C++ •
  34. 34. • Static Interface • • • Modular Interface • Python Octave • • libshogun • C++ •
  35. 35. Windows Cygwin http://www.shogun-toolbox.org/#releases
  36. 36. Windows Linux (Ubuntu) Cygwin sudo apt-get install shogun http://www.shogun-toolbox.org/#releases
  37. 37. Windows Linux (Ubuntu) Cygwin sudo apt-get install shogun http://www.shogun-toolbox.org/#releases Mac sudo port install shogun
  38. 38. Windows Linux (Ubuntu) Cygwin sudo apt-get install shogun http://www.shogun-toolbox.org/#releases Mac sudo port install shogun OK
  39. 39. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  40. 40. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  41. 41. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  42. 42. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  43. 43. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM libsvm c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  44. 44. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM libsvm c1 C 1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  45. 45. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  46. 46. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  47. 47. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  48. 48. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  49. 49. SVM •• libsvm (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify out.txt
  50. 50. •• libsvm (Cmdline ) set_kernel SIGMOID REAL 50 3 0 (cache, gamma, coeff) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIBSVM libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify out.txt
  51. 51. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  52. 52. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  53. 53. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  54. 54. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  55. 55. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT libsvm c1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  56. 56. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT libsvm c1 C 1 train_classifier save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  57. 57. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  58. 58. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  59. 59. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  60. 60. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify
  61. 61. SVM •• svmlight (Cmdline ) set_kernel GAUSSIAN REAL 10 1.2 (cache, kernel width) set_features TRAIN ../data/fm_train_real.dat set_labels TRAIN ../data/label_train_twoclass.dat new_classifier LIGHT libsvm c1 C 1 train_classifier SVM save_classifier libsvm.model load_classifier libsvm.model LIBSVM set_features TEST ../data/fm_test_real.dat out.txt=classify out.txt
  62. 62. Python • sg ( from sg import sg ) • sg OK • Cmdline set_feature TEST data.dat • Python sg(‘set_feature’, ‘TEST’, ‘data.dat’) http://www.shogun-toolbox.org/doc/static_tutorial.html
  63. 63. • SHOGUN
  64. 64. • SHOGUN •3
  65. 65. • SHOGUN •3 • Static Interface,Modular Interface, libshogun
  66. 66. • SHOGUN •3 • Static Interface,Modular Interface, libshogun • Static Interface
  67. 67. • SHOGUN •3 • Static Interface,Modular Interface, libshogun • Static Interface •
  68. 68. • SHOGUN •3 • Static Interface,Modular Interface, libshogun • Static Interface •
  69. 69. • SHOGUN •3 • Static Interface,Modular Interface, libshogun • Static Interface • Modular Interface
  • toshihikonisimura

    Jun. 24, 2015
  • yoshiakiyamada18

    Mar. 4, 2015
  • TakashiNaito3

    Dec. 10, 2014
  • yuichinitto

    Dec. 7, 2014
  • koromoya

    Jan. 16, 2013
  • TakeshiWatanabe2

    Oct. 21, 2012
  • xiangze

    Oct. 21, 2012
  • oghi57

    Aug. 5, 2012
  • takmin

    Apr. 23, 2011

第1回:はじめてのSHOGUN編

Views

Total views

8,808

On Slideshare

0

From embeds

0

Number of embeds

4,126

Actions

Downloads

0

Shares

0

Comments

0

Likes

9

×