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.

続・わかりやすいパターン認識4章(関西機械学習勉強会発表スライド)

1,407 views

Published on

関西機械学習勉強会_続パタ4章の発表資料です。

Published in: Data & Analytics
  • Be the first to comment

  • Be the first to like this

続・わかりやすいパターン認識4章(関西機械学習勉強会発表スライド)

  1. 1. 4 2017/01/15 Nomoto Eriko
  2. 2. / 3200 ( ) 4 2
  3. 3. , , 3
  4. 4. 4
  5. 5. (3.2) (4.1) (4.2) , 5
  6. 6. 6
  7. 7. 7
  8. 8. 8
  9. 9. 4.1 θ (0<θ<1) 1 . n , n r . , . , θ . 9
  10. 10. θ (4.4,4.5) 10
  11. 11. , . (4.6) . 11
  12. 12. (4.10) 12
  13. 13. θ ( ) → 13 ( 1.3 )
  14. 14. θ (4.11) 14 cf. θ
  15. 15. 15 cf. θ θ (4.11) (4.12)
  16. 16. θ (4.11) (4.12) (4.16,4.17) 16 cf. θ θ
  17. 17. 17 cf. θ θ (4.11) (4.12)
  18. 18. θ . (4.18) (4.19) 18 (4.20)
  19. 19. (4.21) (4.23) 19 (4.22) (4.24) (4.25) α 0
  20. 20. (4.20) (4.26) (4.27) 20
  21. 21. (4.17) (4.18) (4.27) (4.11) 21 (4.28) (´ )/
  22. 22. . ( ) 22
  23. 23. (4.28) (4.29) 23 1
  24. 24. (4.28) (4.29) (4.32) 24
  25. 25. 25
  26. 26. 26
  27. 27. 27
  28. 28. 28
  29. 29. 29
  30. 30. 30
  31. 31. 31
  32. 32. 32
  33. 33. 33
  34. 34. 34
  35. 35. =α-1 =β-1 =α+β-2 35
  36. 36. =α-1 =β-1 =α+β-2 36 10
  37. 37. Be(α, β) . B(α, β) (α+β-2) (α-1) , (β-1) . 37
  38. 38. 38 (4.42)
  39. 39. 39 (4.42)
  40. 40. 40 (4.42)
  41. 41. 41 (4.42) !!
  42. 42. . θ . 42
  43. 43. 4.1 θ . . ex)E[θ], M[θ] M[θ] . ( ) . 43
  44. 44. 44 ( ) 0.5
  45. 45. 0.8 10 . 4.32 . 45
  46. 46. 46
  47. 47. 0.8 10 . 4.32 . n=10 100, 1000 . 47
  48. 48. 48 θ
  49. 49. 49
  50. 50. 50
  51. 51. 0.8 n . Be(3, 9) . n=10, 100, 1000 . 51 0.2
  52. 52. 52
  53. 53. 0.8 n . Be(3, 9) . Be(31, 21) . 53 0.2 0.6
  54. 54. 54
  55. 55. , . 0.2 55 0.2
  56. 56. , . , . 56
  57. 57. … (4.49) … (4.50) 57
  58. 58. … (4.49) … (4.50) (4.51) 58 +
  59. 59. … (4.49) … (4.50) (4.51) 59
  60. 60. … (4.49) … (4.50) (4.51) 60
  61. 61. … (4.49) … (4.50) (4.51) 61
  62. 62. 62
  63. 63. 63
  64. 64. 64
  65. 65. 65 4.2 , m 1 . k . , . n , . , . , (k=1~m) .
  66. 66. : 66
  67. 67. (4.59) . 67
  68. 68. 68
  69. 69. 69
  70. 70. 70
  71. 71. 71 t
  72. 72. 72 t t+1
  73. 73. 73 m-1
  74. 74. 74 θ (4.59)
  75. 75. (4.68) 75 (4.59) (4.67)
  76. 76. 76
  77. 77. 77
  78. 78. 78
  79. 79. 79
  80. 80. 80
  81. 81. 81
  82. 82. 82
  83. 83. 83
  84. 84. (4.68) 84 (4.59) (4.67) (4.69)
  85. 85. 85 . m=2
  86. 86. 86 Dir(4, 4, 4)
  87. 87. 87 Dir(1, 1, 1) Dir(4, 4, 4) Dir(8, 8, 8) Dir(8, 2, 2) Dir(2, 8, 2) Dir(2, 4, 6)
  88. 88. 88
  89. 89. 89
  90. 90. 90
  91. 91. 92 m(_ _)m
  92. 92. 93
  93. 93. 94
  94. 94. 95
  95. 95. 96
  96. 96. (4.59) (4.79) 97
  97. 97. cf. 98
  98. 98. https://github.com/NomotoEriko/zokupata 99

×