Forecasting Oncology Demand Trends with Boosting-Based Bayesian Conjugate Models

2026-05-06

AI Summary

This study addresses the challenge of inaccurate forecasting of oncology outpatient demand by proposing a Bayesian prediction framework enhanced with a residual-driven boosting mechanism. The approach innovatively integrates a boosting strategy into a Poisson–Gamma conjugate structure, combining a Gamma–lognormal prior, Poisson process modeling, and Bayesian online updating to effectively capture both short-term fluctuations and long-term trends while preserving analytical tractability. Empirical evaluation on real-world data from Cariri, Brazil, demonstrates that the proposed model achieves up to a 38.25% improvement in trend-direction prediction accuracy over the best-performing baseline, significantly outperforming established methods such as ARIMA, LSTM, and XGBoost.

Abstract

Accurate trend forecasting in healthcare time series is essential for planning and resource allocation. This paper proposes a Bayesian framework for predicting oncology demand trends, modeling weekly appointments as a Poisson process with a Gamma prior to the demand rate. To enhance adaptability and capture persistent directional patterns, we incorporate a residual-based boosting mechanism grounded in a Gamma-Log-Normal conjugate structure. This boosting approach allows the model to track both short- and long-term trend shifts while maintaining the analytical tractability of conjugate Bayesian updating. The methodology was evaluated on real oncology service data from Cariri, Ceara, Brazil, and compared against established baselines, including linear regression, ARIMA, naive forecasting, LSTM neural networks, and XGBoost. Results showed that the proposed model outperforms competing methods in trend detection accuracy, with gains in terms of percentage of correct direction of 38.25% in relation to the second best approach in some cases.

Problem

• oncology demand forecasting
• healthcare time series
• trend prediction
• resource allocation
• Bayesian modeling

Innovation

• Bayesian conjugate models
• boosting mechanism
• Poisson-Gamma process
• trend forecasting
• healthcare time series