DEVELOPMENT OF GHOST WORKERS DETECTION SYSTEM USING BIOMETRIC BASED CONVOLUTIONAL NEURAL NETWORK

Subject Area: COMPUTER SCIENCE

This paper presents the development of ghost workers detection system using biometric based convolutional neural network. The study started with literature review of related works which revealed the impact of ghost worker syndrome on the global economy in general. The study proposes to solve this problem using machine learning based biometric technology. The methods used were data collection, data acquisition, computer vision, face detection, Convolutional Neural Network (CNN), face recognition and results. The methods were designed using structural and mathematical approaches and then implemented with MATLAB. The CNN algorithm for facial recognition was evaluated and the result showed accuracy of 99.28%.