Identification of survival models that are prognostic across cohorts and stable regarding variable selection with methods of model-based optimization

通过基于模型的优化方法，识别在各队列中具有预后性且在变量选择方面稳定的生存模型

• 批准号: 289820878
• 负责人: Professor Dr. Bernd Bischl
• 金额: --
• 依托单位: Lehrstuhl Statistische Methoden in der Genetik und Chemometrie
• 依托单位国家: 德国
• 项目类别: Research Grants
• 财政年份: 2016
• 资助国家: 德国
• 起止时间: 2015-12-31 至 2019-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/289820878?language=en
• 关键词: Identification; survival; models; prognostic; across

The goal is the development of new statistical methods for the identification of survival models that are prognostic across cohorts and stable regarding variable selection. There is a large number of prediction methods for survival data based on clinical and high-dimensional genetic data, and the best model often depends on the patient cohort used. With modern methods of model-based optimization the best models for individual cohorts can be determined efficiently and with significantly reduced run times. The resulting models have two drawbacks. First, they are specialized for the learning cohort and thus often lack of high prediction accuracy on independent cohorts. Second, due to the inherent redundancy in the genetic measurements often different variables are selected on different cohorts despite the same biomedical question. For a better generalizability in this project methods are developed for the identification of models that are at the same time prognostic across cohorts and stable regarding variable selection. A model identified in such a multi-objective approach must then be compared with the cohort-specialized models in order to evaluate the loss in prediction accuracy due to the additional stability criteria. The reproducibility can be significantly improved by integrating public open access experiment databases, since comparison studies of prediction methods then can be extended collaboratively and transparently. The result of this work will be models that are prognostically relevant with a stable variable selection that allows a biological interpretation of the genetic features.

我们的目标是开发新的统计方法，用于识别跨队列预后和变量选择稳定的生存模型。基于临床和高维遗传数据的生存数据预测方法有很多，最佳模型往往取决于所使用的患者队列。使用基于模型的优化的现代方法，可以有效地确定单个队列的最佳模型，并显着减少运行时间。由此产生的模型有两个缺点。首先，它们专门用于学习队列，因此通常缺乏对独立队列的高预测准确性。其次，由于遗传测量中固有的冗余，尽管有相同的生物医学问题，但在不同的队列中通常选择不同的变量。为了更好地概括本项目的方法，开发了用于识别模型的方法，这些模型同时具有跨队列的预测性和关于变量选择的稳定性。然后，必须将在这种多目标方法中确定的模型与队列专用模型进行比较，以评估由于额外的稳定性标准而导致的预测准确性损失。通过整合公共开放获取实验数据库，可以显着提高重现性，因为预测方法的比较研究可以协同和透明地扩展。这项工作的结果将是一个稳定的变量选择，允许遗传特征的生物学解释的模型，是相关的。

项目成果

ReinBo: Machine Learning Pipeline Conditional Hierarchy Search and Configuration with Bayesian Optimization Embedded Reinforcement Learning

• DOI: 10.1007/978-3-030-43823-4_7
• 发表时间: 2019-09
• 作者: Xudong Sun;Jiali Lin;B. Bischl

High Dimensional Restrictive Federated Model Selection with multi-objective Bayesian Optimization over shifted distributions

• DOI: 10.1007/978-3-030-29516-5_48
• 发表时间: 2019-02
• 期刊: ArXiv
• 作者: Xudong Sun;Andrea Bommert;Florian Pfisterer;J. Rahnenführer;Michel Lang;B. Bischl

First Investigations on Noisy Model-Based Multi-objective Optimization

• DOI: 10.1007/978-3-319-54157-0_21
• 发表时间: 2017-03
• 作者: Daniel Horn;Melanie Dagge;Xudong Sun;B. Bischl

Benchmark for filter methods for feature selection in high-dimensional classification data

• DOI: 10.1016/j.csda.2019.106839
• 发表时间: 2020-03-01
• 期刊: COMPUTATIONAL STATISTICS & DATA ANALYSIS
• 影响因子: 1.8
• 作者: Bommert, Andrea;Sun, Xudong;Lang, Michel
• 通讯作者: Lang, Michel