Abstract

The objective of this paper is to propose an efficient regression algorithm of survival analysis - SurvivalBoost.. This algorithm is based on Random Survival Forests (RSF) and XGBoost. By combining the Elastic-Net penalty type Cox proportional hazards regression model with XGBoost optimal algorithm, our algorithm is more suitable for survival analysis. The performance of the proposed algorithm is compared with the Cox proportional hazards regression model, XGBoost, CoxBoost, RSF and Gradient Boosting Desicion Tree-based survival regression model on 4 simulated datasets and 4 real survival datasets. The results illustrated the superiority of the proposed algorithm.

摘要

本文的目的是提出一种有效的生存分析回归算法——SurvivalBoost。该算法基于随机生存森林（RSF）和XGBoost。通过将Elastic-Net惩罚型Cox比例风险回归模型与XGBoost优化算法相结合，我们的算法更适合生存分析。在 4 个模拟数据集和 4 个真实生存数据集上，将所提出算法的性能与 Cox 比例风险回归模型、XGBoost、CoxBoost、RSF 和基于梯度提升决策树的生存回归模型进行了比较。结果说明了该算法的优越性。