| dc.description.abstract | Transfer of knowledge in most institutions of higher learning is done through lecturing where students are expected to attend classes and a lecturer dictates notes and share the contents of the lesson with the students. Overtime, class non-attendance has been frequent in most institutions and has become a concern since a negative relationship exists between attendance and the overall performance. Non-attendance makes a boring lecture environment, affects class success as well as faculty morale as learning declines and academic standards are compromised. Current methods used to manage class attendance which are mostly use of sign sheets have been abused by students signing for their counterparts. Further, they are known to be misplaced and represented later and therefore not presenting the actual status. Data analyses using the current method is cumbersome due to their manual nature and linking with timetable and classroom is also a big issue. This study proposes use of fingerprint biometric identification for class attendance. The students undertaking a given course are registered first then their fingerprints are captured and saved in a database. This is later used to identify students during class attendance for a given lesson. The developed system allows only registered students and it also helps to shows a student percentage attendance for a lesson thus helps easier manage and monitor the attendance policy. An experiment was conducted on the system accuracy, performance and effectiveness which showed the system is very accurate since the Weighted Error Rate from the sample was zero, the True Acceptance Rate was one and the Crossover Error Rate was at zero percent. Performance accuracy was achieved from the False Acceptance Rate and False Rejection Rate being at 0. The system was also able to allow all the users from the sample taken hence it is highly inclusive. | en_US |
