EEG spectrogram classification employing ANN for IQ application

The term intelligence is associated in many areas such as linguistic, mathematical, music and art. In this paper, Intelligence Quotient (IQ) is measured using Electroencephalogram (EEG) from the human brain. The EEG signals are then used to form the spectrogram images, from which a large data of Gray Level Co-occurrence Matrix (GLCM) texture features were extracted. Then, Principal Component Analysis (PCA) is used to reduce the big matrix, and is followed with the classification of the EEG spectrogram image in IQ application using ANN algorithm. The results are then validated based on the concept of Raven´s Standard Progressive Matrices (RPM) IQ test. The results showed that the ANN is able to classify the EEG spectrogram image with 88.89% accuracy and 0.0633 MSE.