Master’s Degree Awarded to Researcher Ahlam Yafaa in Statistics and Information

Researcher Ahlam Abdulwahid Fareh Yafaa was awarded a Master’s Degree in Statistics and Information with an average of Excellent and a grade of (95%) for her thesis titled: An Actuarial Model for Forecasting the Volume of Motor Insurance Claims Using Individual and Hybrid ARIMA and ANN Approaches: An Applied Study on Mareb Yemen Insurance Company, which was submitted to the Faculty of Commerce and Economics – Sana’a University. The MA defense was held on Tuesday, July 14, 2026‎‎.
The MA Viva-voce Committee, which was formed based on a resolution issued by the Graduate Studies and Scientific Research Council, consisted of the following:
• Prof. Abdullah Abdu Mudahesh Ali – Taiz University – External Examiner and Committee Chair.
• Assoc. Prof. Mansour Ahmed Yahya Al-Ameri – Sana’a University – Principal Supervisor.
• Assoc. Prof. Fouad Abdu Ismail Al-Mekhlafi – Sana’a University – Internal Examiner and Committee Member.
The study aimed to develop a model for forecasting the monthly volume of insurance claims in the motor insurance branch by comparing the performance of linear statistical models, represented by the ARIMA model, with nonlinear neural network models (LSTM, GRU, and MLP), as well as hybrid models (ARIMA-ANN and ANN-ARIMA), in terms of their ability to represent and forecast the data accurately.
The study also aimed to compare the performance of feedforward and recurrent neural networks in representing nonlinear temporal relationships and to analyze the effect of the ordering of components in hybrid models on forecasting accuracy.
The findings showed that the LSTM neural network model outperformed the other models, achieving the best forecasting performance, followed by the hybrid LSTM-ARIMA model. The performance of the hybrid models varied according to the ordering of their components, with the LSTM-ARIMA model achieving higher forecasting accuracy than the ARIMA-LSTM model.
The study recommended that insurance companies employ deep-learning techniques in reserve management, pricing, and stress testing.
The thesis defense was attended by faculty members, researchers, students, and interested participants as well as the researcher’s colleagues and family members.