The Brain-Computer Interface (BCI) has significant potential for use in cognitive computation within the field of human-machine interaction. The study was performed to identify the best electrode leads to measure the cognitive activity. Intelligent algorithms were deployed to identify the best electrode leads. EEG signal dataset was obtained from a standard dataset where the signal was acquired from various lead positions. The features from several leads were extracted and contributing features were fed into the Machine Learning classifiers (ML) such as Support Vector Machine (SVM), K-nearest neighbour (KNN), Artificial Neural Network (ANN), and Random Forest (RF). Performance metrics were computed for all the classifiers. SVM achieved 93.71% accuracy for the F3 electrode and 100% accuracy for the F4, P3, and P4 electrodes, while the ANN model achieved 100% accuracy for the F3 electrode and 93.71% for the F4, P3, and P4 electrodes. F3, F4, P3, and P4 are the best suitable leads for cognitive analysis using EEG.