A master's student (Zainab Khamis Matar) in the Department of Computer Science / University of Technology had a discussion about her tagged thesis (Detecting Fraud in Business Intelligence Using Predictive Analytics) This thesis aims to improve the accuracy and performance of the MPSO algorithm. The proposed solution was developed when two main algorithms were combined, SVM and MPSO. The SVM algorithm is used as a function to calculate the accuracy of the learning as MPSO requires the function of calculating the location precision to find the best solution and location of the swarms. Moreover, MPSO algorithm re-transmits the best mode as feedback on SVM algorithm for best accuracy at optimum time. MPSO algorithm issues have been addressed and are limited to two main parts: learning room management and calculating precision function in SVM. Therefore, the proposed solution proposes a new learning room management mechanism for MPSO acceleration. Moreover, the operation of the SVM is improved by keeping the main information and the content of the dataset as a constant aggregation ... That was on Wednesday (1/8/2020) and on the discussion room in the section annex building and the discussion committee was composed of (Prof. Dr. Rana Farid Ghani and Prof. Dr. Suhad Mal Allah Kazem) from the University of Technology / Department of Computer Science and (Prof. Ali Makki Sagheer) from Anbar University / Department of Computer Science in the presence of the student's supervisor (Prof. Dr. Abeer Tariq Mawloud).

Download the abstract