A Biometric Model for Examination Screening and Attendance Monitoring in Yaba College of Technology  Rufai M.M, Adigun J. O, N. A. Yekini Department of Computer Technology, Yaba College of Technology. Abstract: Examination malpractices have consistently remained a bane of Nigerian  educational system. A common form of examination malpractices is the deliberate  impersonation of the applicant. Part of the requirement of a credible examination is  that the real applicants wrote the exams. Several steps have been taken to check this  crime unabated. Some of the methods adopted are: the use of Identity card; the  presence of invigilators to identify fake students; the allocation of sitting  arrangement number that determines the hall where the student will write exam and  the need to sign in and out on the attendance sheets. This research work proffers  solution to the problem of student impersonation during exams. A biometric model is  designed to identify every applicant at the point of entry into the examination hall. A  biometric verification exercise is also conducted while the examination is going and at the point of submission of examination papers. The students’ attendance is captured  automatically as their identity is verified on the biometric systems. The Biometric  Access Control Techniques is explained. A model describing its application to  examination screening and attendance monitoring is designed using denotational  mathematics. Keywords : Biometrics ; Examination Malpractices ; Attendance Monitoring. Swap Training: A Genetic Algorithm Based Feature Selection Method Applied on Face Recognition System Ashkan Parsi, Ali Doostmohammadi Department of Computer Engineering and IT, Shahrood University of Technology, Iran. Mehrdad Salehi, Computer Aided Medical Procedures Technische Universität München, Germany.  Abstract: This paper presents a new feature selection method by modifying fitness  function of genetic algorithm. Our implementation environment is a face recognition  system which uses genetic algorithm for feature selection and k-Nearest Neighbor as a  classifier together with our proposed Swap Training. In each iteration of genetic  algorithm for assessment of one specific chromosome, swaps training switch the  training and test data with each other. By using this method, genetic algorithm does  not quickly converge to local minimums and final recognition rate will be enhanced.  Obtained results from implementing the proposed technique on Yale Face database  show performance improvement of genetic algorithm in selecting proper features.  Keywords :   face recognition ; feature selection ; principal component analysis ;