We use cookies to ensure you get the best experience on our website
Home Explore Help News
Register Sign In
juliankosciessa
/
eegmp_preproc
DOI
10.12751/g-node.6nhr1y
1
0
Fork 1
Files Issues 0 Pull Requests 0 Wiki
Tree: 749a5edc7e
Branches Tags
eegmp_preproc
/
data
/
outputs
/
history
/
log_sub-033_ica1_16-Nov-2021.txt

log_sub-033_ica1_16-Nov-2021.txt 4.2 KB
History Download

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172 
  1. 

  2. Input data size [72,1094500] = 72 channels, 1094500 frames
    3. 

  3. Finding 72 ICA components using extended ICA.
    4. 

  4. Kurtosis will be calculated initially every 1 blocks using 6000 data points.
    5. 

  5. Decomposing 211 frames per ICA weight ((5184)^2 = 1094500 weights, Initial learning rate will be 0.001, block size 70.
    6. 

  6. Learning rate will be multiplied by 0.98 whenever angledelta >= 60 deg.
    7. 

  7. More than 32 channels: default stopping weight change 1E-7
    8. 

  8. Training will end when wchange < 1e-07 or after 512 steps.
    9. 

  9. Online bias adjustment will be used.
    10. 

  10. Removing mean of each channel ...
    11. 

  11. Final training data range: -1.90902 to 2.37523
    12. 

  12. Computing the sphering matrix...
    13. 

  13. Starting weights are the identity matrix ...
    14. 

  14. Sphering the data ...
    15. 

  15. Beginning ICA training ... first training step may be slow ...
    16. 

  16. Lowering learning rate to 0.0009 and starting again.
    17. 

  17. Lowering learning rate to 0.00081 and starting again.
    18. 

  18. Lowering learning rate to 0.000729 and starting again.
    19. 

  19. Lowering learning rate to 0.0006561 and starting again.
    20. 

  20. Lowering learning rate to 0.00059049 and starting again.
    21. 

  21. Lowering learning rate to 0.000531441 and starting again.
    22. 

  22. Lowering learning rate to 0.000478297 and starting again.
    23. 

  23. Lowering learning rate to 0.000430467 and starting again.
    24. 

  24. Lowering learning rate to 0.00038742 and starting again.
    25. 

  25. Lowering learning rate to 0.000348678 and starting again.
    26. 

  26. Lowering learning rate to 0.000313811 and starting again.
    27. 

  27. Lowering learning rate to 0.00028243 and starting again.
    28. 

  28. Lowering learning rate to 0.000254187 and starting again.
    29. 

  29. Lowering learning rate to 0.000228768 and starting again.
    30. 

  30. Lowering learning rate to 0.000205891 and starting again.
    31. 

  31. Lowering learning rate to 0.000185302 and starting again.
    32. 

  32. Lowering learning rate to 0.000166772 and starting again.
    33. 

  33. Lowering learning rate to 0.000150095 and starting again.
    34. 

  34. Lowering learning rate to 0.000135085 and starting again.
    35. 

  35. Lowering learning rate to 0.000121577 and starting again.
    36. 

  36. Lowering learning rate to 0.000109419 and starting again.
    37. 

  37. Lowering learning rate to 9.84771e-05 and starting again.
    38. 

  38. Lowering learning rate to 8.86294e-05 and starting again.
    39. 

  39. Lowering learning rate to 7.97664e-05 and starting again.
    40. 

  40. Lowering learning rate to 7.17898e-05 and starting again.
    41. 

  41. Lowering learning rate to 6.46108e-05 and starting again.
    42. 

  42. Lowering learning rate to 5.81497e-05 and starting again.
    43. 

  43. Lowering learning rate to 5.23348e-05 and starting again.
    44. 

  44. Lowering learning rate to 4.71013e-05 and starting again.
    45. 

  45. Lowering learning rate to 4.23912e-05 and starting again.
    46. 

  46. Lowering learning rate to 3.8152e-05 and starting again.
    47. 

  47. Lowering learning rate to 3.43368e-05 and starting again.
    48. 

  48. Lowering learning rate to 3.09032e-05 and starting again.
    49. 

  49. Lowering learning rate to 2.78128e-05 and starting again.
    50. 

  50. Lowering learning rate to 2.50316e-05 and starting again.
    51. 

  51. Lowering learning rate to 2.25284e-05 and starting again.
    52. 

  52. Lowering learning rate to 2.02756e-05 and starting again.
    53. 

  53. Lowering learning rate to 1.8248e-05 and starting again.
    54. 

  54. Lowering learning rate to 1.64232e-05 and starting again.
    55. 

  55. step 1 - lrate 0.000016, wchange 3978505395368447.00000000, angledelta  0.0 deg
    56. 

  56. Lowering learning rate to 1.18247e-05 and starting again.
    57. 

  57. step 1 - lrate 0.000012, wchange 172280250664.36325073, angledelta  0.0 deg
    58. 

  58. Lowering learning rate to 8.51379e-06 and starting again.
    59. 

  59. step 1 - lrate 0.000009, wchange 123296301.21217768, angledelta  0.0 deg
    60. 

  60. step 2 - lrate 0.000009, wchange 14726552920815772.00000000, angledelta  0.0 deg
    61. 

  61. Lowering learning rate to 6.12993e-06 and starting again.
    62. 

  62. step 1 - lrate 0.000006, wchange 668315.51372326, angledelta  0.0 deg
    63. 

  63. step 2 - lrate 0.000006, wchange 447703037287.44696045, angledelta  0.0 deg
    64. 

  64. step 3 - lrate 0.000005, wchange 19518800702113988.00000000, angledelta  0.4 deg
    65. 

  65. Lowering learning rate to 3.53084e-06 and starting again.
    66. 

  66. step 1 - lrate 0.000004, wchange 2202.51289054, angledelta  0.0 deg
    67. 

  67. step 2 - lrate 0.000004, wchange 4940482.67753164, angledelta  0.0 deg
    68. 

  68. step 3 - lrate 0.000004, wchange 11216713553.37013054, angledelta  6.3 deg
    69. 

  69. step 4 - lrate 0.000003, wchange 5162077916226.53710938, angledelta  6.3 deg
    70. 

  70. step 5 - lrate 0.000002, wchange 666647674977103.00000000, angledelta  6.3 deg
    71. 

  71. step 6 - lrate 0.000002, wchange 28035604311622132.00000000, angledelta  6.3 deg
    72. 

  72. Lowering learning rate to 1.30161e-06 and starting again.
    73.