Performance Analysis of Machine Learning Algorithms for the Detection of Epilepsy

Manual detection of epileptic seizure is a critical task by monitoring electroencephalogram (EEG) signals. It is also a very laborious job and requires highly trained experts. Machine learning based approaches are widely used to solve the above said problem due to its capability to detect seizure by processing long length of data sequences and it also provides a high detection rate. This study provides a machine learning based approach in the spatial domain to find the epileptic seizure patient. We have used three different models and two various dimensionality reduction techniques; Principal Component Analysis, and Linear Discriminant Analysis (LDA). After dimensionality reduction features are fed to the different classifiers, Support Vector Machine (SVM), Logistic regression (LR), and K-Nearest Neighbors (K-NN) to diagnose the epilepsy automatically. This proposed method yielded a maximum accuracy of $\mathbf{9 6 \%}$ for the combination of PCA with SVM.