Data Science and Machine Learning
-
Steel Defect Detection (Kaggle competition)
A multi-class segmentation problem to detect and mask various types of defects in images of steel sheets. There are 4 classes of steel defects. Each image may have no defects, a defect of a single class, or defects of multiple classes.
This was my first participation in a Kaggle competition, top 15% on the Private Leaderboard was achieved. -
ICT Market Study for Saudi Arabia
Participated as a key member in a team that successfully delivered a $3 Million project on conducting ICT Market Study in KSA. Outcomes are reported at International bodies like ITU and G20 Summit and utilized to devise crucial recommendations to CITC, Riyadh. Tasks included: benchmarking study, questionnaire design (using Qualtrics, Key Survey), sampling strategy, deployment plan, data collection and processing, data analysis and reporting, and design of a BI dashboard using Qlik.