EFA JOSEPH EKA

Data Analyst with experience maintaining accurate records, supporting administrative processes, and providing user support across education, healthcare, and logistics environments. Skilled in data entry, data analysis, verification, and system updates while working to defined procedures and deadlines. Experienced supporting teams and service users through clear communication and accurate information management.

Credit Card Fraud Detection | Python, Scikit-learn

• Designed and evaluated machine learning models for fraud detection in severely imbalanced dataset
• Applied class-weighting, SMOTE, and probability threshold tuning to optimise recall-precision trade-offs
• Achieved 94% fraud precision and 82% recall using Random Forest with minimal false positives
• Conducted model comparison and feature importance analysis to support model interpretability

Traffic Accident Data Analysis & Forecasting (Python, SQL, Machine Learning)

• Queried and merged accident, vehicle, and casualty tables in SQLite for integrated analysis
• Identified peak accident times (8–9 AM, 5–6 PM) and higher weekday frequency, especially Monday and Friday
• Examined pedestrian accidents, finding peaks during school and work travel hours
• Performed clustering to identify urban accident prone zones in Kingston upon Hull
• Built Linear Regression and KNN models to forecast weekly accident counts across policing areas
• Forecasted daily accident trends for Hull’s top three high incident LSOAs

• IBM Cybersecurity Analyst Professional Certificate (Completed)
• Security Operations & SIEM
• Incident Response & Digital Forensics
• Network, Cloud & Endpoint Security
• Vulnerability Management
• Threat Hunting & Cryptography