Liangjie Lu — Data Scientist
Statistical rigor • Production-grade analytics • Clear business impact
Experience
7+ years Python/R
SAS certified
Credibility
Fudan (B.S.) + UC Davis (M.S. Stats)
GPA 3.86
Impact
Healthcare, e‑commerce, env. analytics
end‑to‑end delivery
Selected Projects
Evidence-driven case studies with clear business impact.
Clinical Risk Prediction
Healthcare
Cleaned 300k+ EHRs; engineered features; built interpretable risk model.
- Logistic Reg + XGBoost + RF
- SHAP explanations for clinicians
RSASPythonXGBoostSHAP
Impact: AUC ↑ 0.84 → 0.90; PPV +12% at fixed recall
Stormwater GSI Policy Analytics
Public Policy
Regression linking municipal capacity to GSI-supportive policies.
- Data wrangling
- Statistical inference
- Reporting
RStataQuarto
Impact: Explained 62% variance; reproducible report
E‑commerce FM Segmentation
Retail
RFM/FM customer segmentation with dashboard for marketing actions.
- Cohort analysis
- Churn signals
- CLV sketch
PythonPandasPlotly
Impact: Uplift +9% CTR in campaign
Services
From exploratory analysis to productionized models.
Exploratory & Data Audit
End‑to‑end profiling and data quality report.
- Schema & type detection
- Missingness & outliers
- Bias checks
Modeling & Evaluation
From baseline to tuned models with sound validation.
- CV strategy, leakage control
- Interpretability (SHAP/ICE)
- Error analysis
Delivery & Reporting
Actionable dashboards and business narratives.
- Plotly/ECharts dashboards
- Notebooks & slides
- Handover & docs