Feature preprocessing improves Support Vector Machine accuracy for seizure detection in neonatal EEG.
Guy BogaartsErik D. GommerJos P. H. ReulenWerner H. MessDanny M. W. HilkmanVivianne van Kranen-MastenbroekPublished in: IWBBIO (2013)
Keyphrases
- preprocessing
- video recordings
- support vector machine
- epileptic seizures
- detection accuracy
- eeg signals
- detection rate
- feature vectors
- preprocessing step
- detection algorithm
- support vector machine classifier
- image features
- multi class
- high accuracy
- classification accuracy
- false positives
- generalization ability
- detection method
- false alarms
- feature selection
- object detection
- automatic detection
- false positive rate
- brain activity
- machine learning
- face detection
- video data
- anomaly detection
- feature set
- support vector
- feature extraction
- motor imagery
- preprocessing stage
- prediction accuracy
- svm classification
- optimal parameters
- selected features
- svm classifier
- training data
- computer vision