EEG analysis for working memory modeling in face recognition task

The paper proposes a new approach to mode human working memory using EEG-induced fuzzy associativi memory and attempts to recover the encoded memor information from partially supplied instances during memor recall tasks. Experiments are performed to obtain EEG feature from the temporal lobe, representing working memory input an(pre-frontal lobe, representing the working memory outpu during memory encoding process of unknown peoples' face. Thi fuzzy associative memory is built up with these input-outpu features of the working memory, acquired for multiple instances During the recall cycle, we use a notion of fuzzy inversi formulation to recall the input EEG instances from the supplie output instances of the working memory using fuzzy associativ memory, when the subject is asked to remember the original faci from its part, and to our great surprise the model produced inpu instances match with actual brain-generated input instances wit small error. This signifies the importance of the propose(working memory model and inverse formulation in memor retrieval process. Experiments undertaken reveal that the erro metric could successfully be used to diagnose two peopl suffering from Parkinson and three from the early Alzheimer' diseases among a total population of 50 healthy plus 5 brain diseased people.