12345678910111213141516 |
- Input data size [72,1059000] = 72 channels, 1059000 frames
- Finding 72 ICA components using extended ICA.
- Kurtosis will be calculated initially every 1 blocks using 6000 data points.
- Decomposing 204 frames per ICA weight ((5184)^2 = 1059000 weights, Initial learning rate will be 0.001, block size 70.
- Learning rate will be multiplied by 0.98 whenever angledelta >= 60 deg.
- More than 32 channels: default stopping weight change 1E-7
- Training will end when wchange < 1e-07 or after 512 steps.
- Online bias adjustment will be used.
- Removing mean of each channel ...
- Final training data range: -4.049 to 3.68661
- Computing the sphering matrix...
- Starting weights are the identity matrix ...
- Sphering the data ...
- Beginning ICA training ... first training step may be slow ...
- Data has rank 71. Cannot compute 72 components.
|