Advertisement
nomadcrypto

NBA spread model training, validation and backtest results

Nov 29th, 2018
407
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 4.21 KB | None | 0 0
  1. =====================================================
  2. Fold #1
  3. training score 0.651458653213424
  4. testing score 0.6412280701754386
  5. validation score 0.6292
  6. precision recall f1-score support
  7.  
  8. 0 0.65 0.67 0.66 1177
  9. 1 0.63 0.61 0.62 1103
  10.  
  11. micro avg 0.64 0.64 0.64 2280
  12. macro avg 0.64 0.64 0.64 2280
  13. weighted avg 0.64 0.64 0.64 2280
  14.  
  15. =====================================================
  16. =====================================================
  17. Fold #2
  18. training score 0.6760254441763545
  19. testing score 0.6333333333333333
  20. validation score 0.6292
  21. precision recall f1-score support
  22.  
  23. 0 0.62 0.76 0.68 1177
  24. 1 0.66 0.50 0.57 1103
  25.  
  26. micro avg 0.63 0.63 0.63 2280
  27. macro avg 0.64 0.63 0.62 2280
  28. weighted avg 0.64 0.63 0.63 2280
  29.  
  30. =====================================================
  31. =====================================================
  32. Fold #3
  33. training score 0.6613292388681729
  34. testing score 0.6285087719298246
  35. validation score 0.6208
  36. precision recall f1-score support
  37.  
  38. 0 0.62 0.71 0.67 1177
  39. 1 0.64 0.54 0.58 1103
  40.  
  41. micro avg 0.63 0.63 0.63 2280
  42. macro avg 0.63 0.63 0.62 2280
  43. weighted avg 0.63 0.63 0.63 2280
  44.  
  45. =====================================================
  46. =====================================================
  47. Fold #4
  48. training score 0.6614760390393684
  49. testing score 0.6476524791575252
  50. validation score 0.6336
  51. precision recall f1-score support
  52.  
  53. 0 0.65 0.69 0.67 1176
  54. 1 0.64 0.61 0.62 1103
  55.  
  56. micro avg 0.65 0.65 0.65 2279
  57. macro avg 0.65 0.65 0.65 2279
  58. weighted avg 0.65 0.65 0.65 2279
  59.  
  60. =====================================================
  61. =====================================================
  62. Fold #5
  63. training score 0.6648755345980919
  64. testing score 0.6524791575252303
  65. validation score 0.6392
  66. precision recall f1-score support
  67.  
  68. 0 0.65 0.70 0.68 1176
  69. 1 0.65 0.60 0.62 1103
  70.  
  71. micro avg 0.65 0.65 0.65 2279
  72. macro avg 0.65 0.65 0.65 2279
  73. weighted avg 0.65 0.65 0.65 2279
  74.  
  75. =====================================================
  76. kfold train scores mean 0.6630329819790823
  77. kfold train scores [0.651458653213424, 0.6760254441763545, 0.6613292388681729, 0.6614760390393684, 0.6648755345980919]
  78. kfold test score mean 0.6406403624242704
  79. kfold test scores [0.6412280701754386, 0.6333333333333333, 0.6285087719298246, 0.6476524791575252, 0.6524791575252303]
  80. validation scores mean 0.6304000000000001
  81. validation scores [0.6292, 0.6292, 0.6208, 0.6336, 0.6392]
  82.  
  83. Full data results:
  84. train score 0.6706832007487131
  85. test score 0.655438596491228
  86. validation score 0.6384
  87.  
  88. validation brier score 0.30502041562367965
  89. book brier score 0.251380321448
  90. test report:
  91. precision recall f1-score support
  92.  
  93. 0 0.64 0.72 0.68 1452
  94. 1 0.67 0.58 0.62 1398
  95.  
  96. micro avg 0.66 0.66 0.66 2850
  97. macro avg 0.66 0.65 0.65 2850
  98. weighted avg 0.66 0.66 0.65 2850
  99.  
  100. validation report:
  101. precision recall f1-score support
  102.  
  103. 0 0.64 0.67 0.65 1281
  104. 1 0.63 0.61 0.62 1219
  105.  
  106. micro avg 0.64 0.64 0.64 2500
  107. macro avg 0.64 0.64 0.64 2500
  108. weighted avg 0.64 0.64 0.64 2500
  109.  
  110. validation back test results:
  111. starting bankroll $500.00 ending bankroll $2508472.14
  112. wins 1596.0 losses 904.0 win percent 0.6384
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement