Zahan, FahmidaMahmud, NusratNawshin, AmannaAhmed, Ferdousi2015-07-022015-07-022015-04ID 10301012ID 10301015ID 10310003ID 10310010http://hdl.handle.net/10361/4211This thesis report is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Electrical and Electronic Engineering, 2015.This thesis verifies that the proposed Kernel mapping based recursive least square algorithm can detect the slightest deviation of anomaly from the norm, monitor and learn underlying pattern between natural and abnormal multivariate medical parameters of a particular critical ICU patient with high detection accuracy and very low rate of false alarm. This online, automated, sequential, real-time intruder detection algorithm is suitable for any instantaneous detection of accidental emergencies without compromising the patient safety and effectiveness of care. It is an elegant, inexpensive solution, independent of complexity, and also a portable and adaptive approach.enElectrical and electronic engineeringThesis