Skip to content
More
Share
Explore
Meetings
Link to page
GPU: NVIDIA GeForce GTX 1080 Ti
Without windowing
Subject 1:
EGNET Model:
Accuracy: 99.94%
Size of training set: 6382
Size of testing set: 1596
classification report
Class
Precision
Recall
F1-score
Support
left
1
1
1
191
reverse
1
1
1
14
right
1
1
1
235
steer
1
1
1
718
stop
1
1
1
438
Total/Average
1
1
1
1,596
There are no rows in this table
DeepConvNet:
Accuracy: 99.87%
Size of training set: 6382
Size of testing set: 1596
Classification Report _
Class
Precision
Recall
F1-score
Support
left
1.00
1.00
1.00
191
reverse
1.00
1.00
1.00
14
right
1.00
1.00
1.00
235
steer
1.00
1.00
1.00
718
stop
1.00
1.00
1.00
438
accuracy
1
macro avg
1.00
1.00
1.00
1,596
weighted avg
1.00
1.00
1.00
1,596
There are no rows in this table
EEGTransformer:
Accuracy: 0.88%
Precision: 0.00
Recall: 0.01
F1-score: 0.00
Size of training set: 6382 (79.99%)
Size of testing set: 1596 (20.01%)
Classification Report __
Class
Precision
Recall
F1-score
Support
left
0.00
0.00
0
191
reverse
0.01
1.00
0.02
14
right
0.00
0.00
0
235
steer
0.00
0.00
0
718
stop
0.00
0.00
0
438
accuracy
0.01
1,596
macro avg
0.00
0.20
0
1,596
weighted avg
0.00
0.01
0
1,596
There are no rows in this table
With Gaussian windowing
Subject 1:
label count
steer 341
stop 233
right 114
left 100
reverse 9
Number of samples before windowing: 7978
Number of samples after windowing: 797
EEGNET Model:
Random splitting training and testing sets:
Accuracy: 99.38%
Size of training set: 637
Size of testing set: 160
Classification Report S1_EEGNnet_random
Class
Precision
Recall
F1-score
Support
left
0.95
1.00
0.98
20
reverse
1.00
1.00
1.00
3
right
1.00
1.00
1.00
22
steer
1.00
0.98
0.99
62
stop
1.00
1.00
1.00
53
Accuracy
0.99
Macro avg
0.99
1.00
0.99
160
Weighted avg
0.99
0.99
0.99
160
There are no rows in this table
Time-based cross-validation (sequential train-test split):
Divide the data into two halves and perform a continuous split (80% training, 20% testing) on each half.
Accuracy: 89.38%
Size of training set: 637
Size of testing set: 160
Classification Report S1_EEGNnet_STTS
Class
Precision
Recall
F1-score
Support
left
1.00
0.64
0.78
11
reverse
0.29
1.00
0.44
2
right
0.81
1.00
0.89
21
steer
0.98
0.91
0.94
96
stop
0.84
0.87
0.85
30
Accuracy
0.89
Macro avg
0.78
0.88
0.78
160
Weighted avg
0.92
0.89
0.90
160
There are no rows in this table
DeepConvNet:
Random splitting training and testing sets:
Size of training set: 637
Size of testing set: 160
Accuracy: 96.88%
Precision: 0.97
Recall: 0.97
F1-score: 0.97
Classification Report S1_DeepConvNet_random
Class
Precision
Recall
F1-score
Support
left
0.95
0.95
0.95
20
reverse
1.00
1.00
1.00
3
right
0.88
1.00
0.94
22
steer
1.00
0.94
0.97
62
stop
0.98
1.00
0.99
53
accuracy
0.97
macro avg
0.96
0.98
0.97
160
weighted avg
0.97
0.97
0.97
160
There are no rows in this table
Time-based cross-validation (sequential train-test split):
Size of training set: 637
Size of testing set: 160
Accuracy: 95.00%
Precision: 0.95
Recall: 0.95
F1-score: 0.95
Classification Report S1_DeepConvNet_STTS
Class
Precision
Recall
F1-score
Support
left
1.00
0.64
0.78
11
reverse
1.00
1.00
1
2
right
0.88
1.00
0.93
21
steer
0.96
0.97
0.96
96
stop
0.97
0.97
0.97
30
Accuracy
0.95
160
Macro avg
0.96
0.91
0.93
160
Weighted avg
0.95
0.95
0.95
160
There are no rows in this table
EEGTransformer:
Random splitting training and testing sets:
Accuracy: 33.12%
Precision: 0.11
Recall: 0.33
F1-score: 0.16
Size of training set: 637
Size of testing set: 160
Classification Report S1_EEGTransformer_random
Class
Precision
Recall
F1-score
Support
left
0.00
0.00
0
20
reverse
0.00
0.00
0
3
right
0.00
0.00
0
22
steer
0.00
0.00
0
62
stop
0.33
1.00
0.5
53
accuracy
0.33
160
macro avg
0.07
0.20
0.1
160
weighted avg
0.11
0.33
0.16
160
There are no rows in this table
Time-based cross-validation (sequential train-test split):
Accuracy: 13.12%
Precision: 0.02
Recall: 0.13
F1-score: 0.03
Size of training set: 637 (79.92%)
Size of testing set: 160 (20.08%)
Classification Report S1_EEGTransformer_STTS
Class
Precision
Recall
F1-score
Support
left
0.00
0.00
0
11
reverse
0.00
0.00
0
2
right
0.13
1.00
0.23
21
steer
0.00
0.00
0
96
stop
0.00
0.00
0
30
accuracy
0.13
160
macro avg
0.03
0.20
0.05
160
weighted avg
0.02
0.13
0.03
160
There are no rows in this table
Subject 2, Session 1:
label count
steer 341
stop 265
left 124
right 94
reverse 11
Number of samples before windowing: 8358
Number of samples after windowing: 835
EEGNET Model:
DeepConvNet:
EEGTransformer:
Subject 2, Session 2:
label count
steer 145
stop 118
left 77
right 67
reverse 0
Number of samples before windowing: 4079
Number of samples after windowing: 407
EEGNET Model:
DeepConvNet:
EEGTransformer:
Subject 3:
label count
steer 313
stop 185
right 155
left 147
reverse 82
Number of samples before windowing: 8827
Number of samples after windowing: 882
EEGNET Model:
DeepConvNet:
EEGTransformer:
Subject 4:
label count
steer 229
stop 151
reverse 46
left 16
right 13
Number of samples before windowing: 4552
Number of samples after windowing: 455
EEGNET Model:
DeepConvNet:
EEGTransformer:
Want to print your doc?
This is not the way.
This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (
CtrlP
) instead.