KC-OverdoseModels2025/XGBoost_Census-Cluster-Rate

XGBoost — Fentanyl Overdose Rate Forecasting

Author: Ansh Gupta

Overview

An XGBoost regression model for forecasting census tract-level fentanyl/opioid overdose rates in King County, Washington. The model integrates American Community Survey (ACS) census data, Group-Based Trajectory Modeling (GBTM) cluster assignments, and historical overdose rates to predict future overdose rates per 100,000 population.

Performance (5-Fold Cross-Validation)

Metric	Value
R²	0.923
MAE	5.79 per 100,000
RMSE	9.47 per 100,000
Spearman ρ	0.929
Top-10 Overlap	90.0%
BPR K=10	100.0%
Top-20 Overlap	100.0%
BPR K=20	100.0%
Top-50 Overlap	86.0%
BPR K=50	98.0%
Mean BPR (K=5–200)	97.6%

Features

Feature	Description
Med_HHD_Inc_Thousands_ACS__	Median household income (thousands)
pct_Renter_Occp_HU_ACS__	% renter-occupied housing units
pct_Vacant_Units_ACS__	% vacant housing units
pct_Not_HS_Grad_ACS__	% without high school diploma
pct_College_ACS__	% with college degree
pct_NH_Blk_alone_ACS__	% non-Hispanic Black population
Pct_No_Health_Ins_CALCULATED_ACS__	% without health insurance
Med_HHD_Inc_Thousands_ACS___Neighbor_Avg	Average median income of 3 nearest neighbor tracts
pct_Vacant_Units_ACS___Neighbor_Avg	Average vacancy rate of 3 nearest neighbor tracts
pct_Renter_Occp_HU_ACS___Neighbor_Avg	Average renter rate of 3 nearest neighbor tracts
Rate_Independent	Prior-year overdose rate (per 100,000)
Cluster_1	GBTM trajectory cluster 1 indicator
Cluster_2	GBTM trajectory cluster 2 indicator
Cluster_3	GBTM trajectory cluster 3 indicator

Hyperparameters

Parameter	Value
n_estimators	500
max_depth	3
learning_rate	0.01
subsample	1.0
colsample_bytree	1.0
objective	reg:squarederror

Validation

• Temporal: Out-of-time forecast (trained on ≤2024 data, predicting 2025 TTM rates)
• Cross-validation: 5-fold CV — every tract predicted by a model never trained on it
• Geographic: Cross-regional transfer validated on Los Angeles County, CA

Training Data

• Region: King County, Washington (463 census tracts)
• Target: Overdose death rate per 100,000 (2025 trailing twelve months)
• Census data: ACS 5-year estimates (2019–2023)

Usage

import xgboost as xgb
import pandas as pd

# Load model
model = xgb.Booster()
model.load_model("xgboost_model.json")

# Prepare features (must match order in features.json)
dtest = xgb.DMatrix(your_features_df)
predictions = model.predict(dtest)

Files

File	Description
xgboost_model.json	Trained XGBoost model (native JSON format)
xgboost_model.pkl	Trained XGBoost model (pickle format)
features.json	Feature names and target variable
hyperparameters.json	Tuned hyperparameters
training_data.csv	Full training dataset (463 tracts)

License

MIT