Aimed to minimize police response time by detecting weapons in a live cctv camera. This model also works on images and videos.
The main motivation of this project is due to the increasing number of school mass shootings in the U.S.
Scrapped 1.6M Reddit posts to create a model that classifies sad/angry posts vs suicidal/depressed posts. Used Logistic regression, mixed and ensemble modeling to get a classification accuracy of 82%. This project was made as an approach to reduce suicide rate by identifying depressed users and offering them help.
Used a variety of regression models to predict house prices at Ames, Iowa between 2006 and 2010. The process includes intensive hyperparameters-tuning and data visualizations. Placed #1 in the Kaggle competition with the most accurate prediction.
Applied Machine learning and data analysis on big data in a 500 hour immersive course
Large-scale data collection automation, machine learning and analysis.
Led Cisco networking classes, mentored fellow students and held open office hours.
Taught cryptography and stenography to high school students. Designed and built five 3D-printed drones.
I left my country in 2012 looking for a safer place to call home. I traveled around Europe, the Middle East, and Asia before settling in the United States.
My curiosity and passion for learning led me to pursue a career in Cyber Security and Data Science, particularly projects related to machine learning and neural networks. When not coding, I enjoy the outdoors by sprinting, hiking, and camping. You can beat me in everything but probably not in sprinting :)
I am eager to expand my knowledge in deep learning and computer vision. If you think we can collaborate, please connect with me!