This repository showcases an end-to-end data analytics and machine learning workflow. I bridge the gap between complex data orchestration and actionable business intelligence, delivering evidence-based clarity for strategic decision-making.
- 🎯 Objective: Translate raw transaction data into retention strategies.
- 💡 Achievement: Processed 500k+ records into an RFM framework and developed a Churn Prediction Model with 90.5% accuracy.
- 📊 Visualization: Executive-level Tableau dashboards serving as a "Source of Truth" for customer health metrics.
- 🎯 Objective: Process high-velocity spatiotemporal data for urban mobility insights.
- 🛠 Tech: Orchestrated a high-performance pipeline for 21.8M+ GPS records using
data.tableanddplyrin R. - 📈 Impact: Derived driver activity patterns and operational mobility metrics from unstructured data.
- 🎯 Objective: Optimize credit risk assessment through algorithmic benchmarking.
- 🔬 Methodology: Compared SVM (AUC: 0.8524), Random Forest, and Naive Bayes; utilized PCA and SOM Neural Networks for dimensionality reduction.
- Languages: R (Advanced), SQL.
- Tools: Tableau.
- BA Frameworks: BPMN 2.0, Gap Analysis, SOA Architecture.
Maintained by Yanhui Ma – Business Analyst & Operations Specialist