bp
45c93dfbbf
Merge branch 'dev' of github.com:Pezz89/QMUL_Final_Project into dev
2017-09-21 10:00:49 +01:00
bp
dfd5fc7465
Added untracked files
2017-09-21 09:58:27 +01:00
bp
62508e570e
Updated readme
2017-09-21 09:56:02 +01:00
bp
b6ad0c0977
Added plotting scripts
2017-08-24 15:39:42 +01:00
bp
444ddd4c52
Final finishing touches
2017-08-22 13:59:54 +01:00
bp
87050198a9
Implemented revised resampling and updated requirements
2017-08-22 11:01:10 +01:00
bp
acae395e81
save
2017-08-22 07:50:44 +01:00
bp
f79d13edbd
added std dev to results
2017-08-21 23:27:11 +01:00
bp
c814922a65
Added references file
2017-08-20 10:17:41 +01:00
bp
9c9a7c60de
Added sensitivity and specificity scoring tables
2017-08-20 00:38:17 +01:00
bp
5a592d8dd8
Added final demonstration scripts
2017-08-19 23:22:15 +01:00
bp
c456774e5b
Altered feature selection CLI arguments again...
2017-08-19 22:57:40 +01:00
bp
3f00875bab
Altered feature selection CLI arguments
2017-08-19 21:55:36 +01:00
bp
feff732605
updated scorer
2017-08-19 17:04:03 +01:00
bp
848011cb9d
Revising scorer - unstable
2017-08-19 16:12:56 +01:00
bp
f0b76002a5
Added PyTables to install script
2017-08-19 13:18:29 +01:00
bp
7c3bbb2bec
Added nojvm to matlab command
2017-08-19 12:35:34 +01:00
bp
bbb38984e2
updated install script
2017-08-19 12:25:54 +01:00
bp
54b218d253
Fixed resamling issue, added demo run script
2017-08-19 12:22:36 +01:00
bp
3af0b11662
updated readme and install scirpt
2017-08-19 10:55:17 +01:00
bp
cce5bb7eee
updated readme and install scirpt
2017-08-19 10:22:34 +01:00
bp
03c6d4b8d0
Updated requirements
2017-08-19 09:02:05 +01:00
bp
a746571b4b
Updated requirements
2017-08-18 23:55:27 +01:00
bp
c2c9235b38
Cleaned up old scripts, begun writing instructions
2017-08-18 23:42:47 +01:00
bp
b6e8edf14c
Finished commenting, added requirements.txt
2017-08-18 21:33:05 +01:00
bp
c6495d8658
Clean up, commenting code
2017-08-18 20:25:09 +01:00
bp
db170ebf37
Commenting code
2017-08-18 20:13:09 +01:00
a4f6308a0d
Added automatic log removal
2017-08-18 13:32:43 +01:00
472b2ae2ef
Added max_features argument for feature selection, begun commenting code
2017-08-18 13:24:55 +01:00
44cf9d2307
Final output now prints selected features
2017-08-18 13:06:58 +01:00
261ca885e4
Formatted results output into tables
2017-08-18 12:53:11 +01:00
4d51d52523
Edited optimization logger output
2017-08-18 12:05:31 +01:00
ad5967cb3f
Added logistic regression to optimization parameters
...
Added logistic regression meta classifier parameters to particle swarm
optimization
2017-08-18 11:56:57 +01:00
a102bc798b
System tested from start to finish successfully
2017-08-16 23:30:17 +01:00
074014c45f
Added stacking classifier as primary classifier
2017-08-16 12:40:36 +01:00
af09bc61dc
Final cross-val score + use of quality labels
...
- Class labels are now loaded using the SQI data provided to set all low
quality signals as unsure
- 10 fold stratified cross-validation now used as final scoring method
2017-08-16 10:51:10 +01:00
4ff30ee4f0
Fixing PCA related problems...
2017-08-13 19:31:53 +01:00
7e65c98d3e
Refactored model building, added AdaBoost
2017-08-12 21:17:15 +01:00
d7ec4c292e
removed modified librosa, project stable
2017-08-11 16:50:19 +01:00
ef6eba711d
Merge commit '5e0dfbf1474802b495b666fc229ac3e36f7b94e3' as 'src/librosa'
2017-08-11 15:51:25 +01:00
5e0dfbf147
Squashed 'src/librosa/' content from commit 74dafbf4
...
git-subtree-dir: src/librosa
git-subtree-split: 74dafbf497d552e77f861c543eee7aec1ff720ec
2017-08-11 15:51:25 +01:00
8f086d79de
minor change
2017-08-11 15:50:25 +01:00
cdfb0f092c
Updated to python 3, implemented MFCC features
...
- Updated to Python 3 in a failed attempt to fix parallelization with
MFCC features
- Global and local MFCC features are now implemented
- Parallelization not currently working but sychronous processing works
2017-07-13 19:18:33 +01:00
5088499ad0
Added new features, significant improvement on results
...
- Added features: signal segment max, signal segment mean, Total power
in time domain, total power in frequency domain, mean MFCCs per segment
selection of wavelet features
- Temporarily disabled PCA
2017-07-13 13:19:54 +01:00
211bca03a6
Implemented wrcoef
2017-07-12 18:17:27 +01:00
fa2ea4a700
Implementing wavelet based features. Unfinished
2017-07-12 07:37:42 +01:00
8b53ff02da
Implementing wavelet based features. Unfinished
2017-07-10 18:25:49 +01:00
81a5e168d6
Implemented Kernel PCA, Sample and Shannon Entropy
...
- Implemented Kernal PCA for feature dimensionality reduction. The best
30 components are now used as initial features
- Implemented Samplke entropy and shannon entropy features
2017-07-10 11:11:25 +01:00
550545b23d
Merge commit '6b396c832bc6a3e0ab823fca31be3d02e6966cd5' as 'src/pyeeg'
2017-07-09 19:08:41 +01:00
6b396c832b
Squashed 'src/pyeeg/' content from commit 34131b7
...
git-subtree-dir: src/pyeeg
git-subtree-split: 34131b78b1a6e24790c610dd80896e1c11404ebc
2017-07-09 19:08:41 +01:00
bp
Sam Perry