College of Computer Science and Mathematics

Discussion of a PhD thesis in the College of Computer Science and Mathematics – Department of Computer Science on the design and construction of an intelligent computer system for classifying facial expressions.A doctoral thesis entitled:
(Behavioral Sense Classification Using Machine Learning Algorithms)
For student Maher Khalaf Hussein Hammadi and under the supervision of Assistant Professor Dr. Fawzia Mahmoud Ramo.The thesis dealt with the design and construction of an intelligent computer system for classifying facial expressions called FERS-DL. This system consists of several intelligent systems to provide various software methods for data processing and obtain the highest classification accuracy.The thesis dealt with the construction of a proposed system of two basic stages, the feature extraction stage and the classification stage, where six methods were proposed in the feature extraction stage: four of them using machine learning algorithms and two methods using deep learning algorithms, and each method combines several intelligent techniques, and to complete the extraction stage Features In order to get good features, after the feature extraction process, the hybrid ant lion algorithm with neural networks was used to select the best features, which helps to increase the classification accuracy.
As for the second stage, which is the classification stage, this stage was designed and implemented in two directions, the first direction using the concept of fine-tuning (Fine-Tune) and the second direction was proposed a new classification algorithm called (EFMNN) based on the method of merging the classifiers, where several neural networks were built to distinguish Seven facial expressions.The thesis aimed to test the system on two standard databases, namely (JAFFE) (CK), and using the concept of Few Shot Learning. Data were used in the training phase with different sizes and percentages ranging from 30 to 70 percent of the total data volume. Very efficient results using the proposed intelligent systemThe discussion committee was chaired by Prof. Dr. Khalil Ibrahim Ahmed, with the membership of Assistant Professor Dr. Iman Saleh Sakban from Babylon University – College of Information Technology, Assistant Professor Dr. Sinan Adnan Diwan from Wasit University – College of Computer Science and Information Technology, Assistant Professor Dr. Nada Nemat Salim, and Assistant Professor Dr. Ghaida Abdel Al-Aziz Majeed, and the membership and supervision of the assistant professor, Dr. Fawzia Mahmoud Ramo.