Edelle Gibben Lumabi

Experienced with data analysis, statistical modeling, and machine learning techniques. Utilizes advanced tools to uncover critical insights and drive strategic decisions. Knowledge of data visualization software to communicate findings effectively.

IBM - Data Science Methodology

• E-Commerce Customer Segmentation (RFM Analysis): I analyzed over 500k UK-based retail transactions using Python and SQL to develop an RFM segmentation model that categorized customers into actionable groups. By implementing advanced CTEs and statistical scoring, I identified that a core "Champions" segment drives 60% of total revenue despite comprising only 15% of the user base. This project delivered data-driven insights into customer retention, specifically highlighting a significant "At Risk" population to inform targeted re-engagement strategies.
• Comparative Evaluation of YOLOv26 Hyperparameter Configurations for Real-Time Urban Vehicle Detection in Philippine Mixed-Traffic Environments: I optimized a YOLOv26 model for real-time vehicle detection in complex Philippine urban environments by engineering a custom dataset of 2,168 annotations. Through a comparative evaluation of three hyperparameter configurations, I achieved a peak mAP50 of 0.8271 and 0.9414 precision using an automated optimization strategy. This project demonstrates high-accuracy detection and generalization across diverse vehicle classes in high-density, mixed-traffic scenarios.
• CardinalStage: Online Booking System for Mapua Performing Organizations: Took part in engineering a full-stack booking system called CardinalStage to centralize and automate the performance management process for Mapua University organizations. Leveraging PHP and SQL, I helped built a dynamic platform that facilitates seamless communication between event organizers and performers through a secure, role-based dashboard.
• CoinBuddy: Gamified Expense Tracker Application: I helped in developing CoinBuddy, a comprehensive financial tracking mobile application built with React Native and Expo, featuring a secure backend integrated through Firebase and Firestore. The project involved implementing complex functionalities such as automated PDF transaction report generation, a QR scanner for profile sharing, and a gamified achievement system to encourage user engagement. I also architected the application's navigation using Expo Router and ensured data privacy by implementing custom Firestore security rules and a dedicated privacy management system.