Mohammad Ariaeimehr


I hold a master’s degree in software engineering and have dedicated three years to research in deep learning. My work focuses on innovative applications of the Transformer model, applying positional encoding to each attention head instead of the traditional input. Additionally, I have explored Federated Learning (FL) to enhance privacy in human activity recognition (HAR). This research has led to a journal paper currently under review. I have completed my master’s degree and am now applying for a Ph.D. program in the United States to further advance my research in AI and machine learning.
I have worked as a system administrator for one of the largest telecom companies in Asia (MCI), gaining extensive experience in operating systems and infrastructure management. I have installed, maintained, and optimized critical systems, ensuring stability and efficiency in a high-demand environment.
I received my M.Sc. in Computer Engineering in 2024. During my master’s degree, I worked on human activity recognition using deep learning. More specifically, I developed an approach to utilize deep learning for human activity recognition in wearable devices. I am particularly interested in applied deep learning and machine learning, especially in healthcare.
Let's talk about fun stuff! I enjoy coffee and tea ☕. I love biking and mountain climbing🏔️🚴 And of course, I follow tech-related topics 💻📱!
Education
Islamic Azad University - Tehran Central Branch
Tehran, Iran
M.Sc. in Software Engineering
- Thesis: Human Activity Recognition with Applied Positional Encoding on Attention Matrix
- Research Focus: Deep Learning, Transformer Model and Federated Learning
Islamic Azad University - Marand Branch
Marand, Iran
BSc. in Electronic Technology Engineering
- Entrepreneurship Project : Modeling low-power comparisons with a supply voltage of 0.5 V and a frequency of 2 MHz
Islamic Azad University - Sofian Branch
Sofian, Iran
Associate Degree in Electronics
Selected Publications
2024
Improved the Performance of Human Activity Recognition Using Transformer Model Based on the Positional Encoding.
Mohammad Ariaeimehr and Reza Ravanmehr
Multimedia Tools and Applications Journal. (Under Review)