Advancement Through Research
Brain-ComputerInterface technology has increased considerable interest, in the automatic detection of states such as facial expression, voice, and physiological signals for developing and ensure the human-computer interaction with an effec-tive way and better way. The brain stimulating signals are detected by the BCI (Brain-computer Interface) which based on the idea of emotional brain comput-er. BCI work can be in three steps first, collecting brain signal second step, in-terpreting the brain signal and third step, output command to connected machine according to the brain signal. the brain wave signals have specific unique electri-cal signals identified with a different individual and they are hard to imitate also. Electroencephalogram (EEG) catches the brain waves as electrical signals and utilized as a part of different applications including therapeutic services and hu-man PC collaboration. In this proposed paper, we discuss local signal pattern transformation based feature extraction technique method for brain signal (EEG). In this paper, two functional feature extraction technique method namely One Dimensional Local Gradient Pattern (1D-LGP) and Local Neighboring De-scriptive Pattern (LNDP) has been discussed for brain signal EEG for machine learning. The EEG raw dataset is provided by and is provided by Neurodynam-ics Laboratory, State University of New York Health Center Brooklyn, New York. The classification performance is evaluated is 10 fold classification tech-nique in machine learning. 1D-LGP and LNDP both feature extraction technique with machine learning classifier achieved 88.90% and 90.93%, classification to EEG brain signal respectively. This study suggests LNDP could be an effective feature extraction technique for classification for EEG signal.
Electroencephalogram (EEG) signals, One Dimensional Local Gradient Pattern (1D-LGP), Local Neighboring Descriptive Pattern (LNDP)
