Comparative Forecasting and Misclassification Analysis Using Health Survey Data

Ermioni Traka

doi:10.3390/ai7040148

2026-04-20

DOI：10.3390/ai7040148

原文链接

查看原文

摘要

En 中文

背景：准确的死亡率预测仍然是公共卫生领域的主要挑战，因为人口、社会经济、行为和医疗因素之间存在复杂的相互作用。这个问题对于识别高风险群体和改进预防性医疗保健策略尤为重要。虽然现有研究展示了强大的预测性能，但它们主要依赖于临床结构化数据，并专注于模型性能。误分类和非典型病例等挑战尚未得到充分探索。方法：本研究使用综合公共使用微数据系列国家健康访谈调查（IPUMS-NHIS）2010年和2015年的数据集（193,765条记录，104个特征），通过比较机器学习方法研究死亡率预测。数据预处理包括特征工程、分类编码和缺失条目移除。采用SMOTE和SMOTE-ENN重采样解决类别不平衡问题，随后进行超参数调整。训练了三种模型——逻辑回归、随机森林和XGBoost——来分类死亡率，优先考虑召回率以确保准确识别死亡病例。结果：结果表明，XGBoost取得了最佳性能（召回率=69%，F1=0.39，AUC=0.92），在敏感性和特异性平衡方面优于其他模型。特征重要性和排列分析强调了年龄、就业状况、自报健康状况和生活方式指标作为关键预测因子。误分类分析结合孤立森林揭示了标准模型未能捕捉到的非典型特征。结论：研究发现强调了XGBoost的有效性，并证明了将异常检测与分类相结合以改进死亡率预测并指导公共卫生规划的价值。

Keyword:

anomaly detection

class imbalance

logistic regression

machine learning

mortality prediction

public health forecasting

random forest

XGBoost

AI总结

对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。

期刊

IF:

论文数:

1.1K

被引数:

941

机构

international hellenic university

学者数:

165

论文数: 102

被引数: 0

university of ioannina

学者数:

588

论文数: 294

被引数: 0

引用论文

Population risk stratification tools and interventions for chronic disease management in primary care: a systematic literature review 基层慢性病管理中的人群风险评估工具和干预措施：一项系统文献综述

health services research

IF3

2025-04-10

OAAI

Golinelli, Davide; Pecoraro, Valentina; Tedesco, Dario; Negro, Antonella; Berti, Elena; Camerlingo, Maria Domenica; Alberghini, Lucia; Rolli, Maurizia; Grilli, Roberto

Deep Patient: An Unsupervised Representation to Predict the Future of Patients from the Electronic Health Records 深度患者: 从电子健康记录中预测患者未来的无监督表示

SCIENTIFIC REPORTS

IF3.9

2016-05-17

1.0K

OAAI

Miotto, Riccardo; Li, Li; Kidd, Brian A.; Dudley, Joel T.

Socioeconomic disparities in healthcare access and implications for all-cause mortality among US adults: a 2000-2019 record linkage study

AMERICAN JOURNAL OF EPIDEMIOLOGY

IF4.8

2024-07-24

PREAI

Gulati, Ishnaa; Kilian, Carolin; Buckley, Charlotte; Mulia, Nina; Probst, Charlotte

SMOTE vs. SMOTEENN: A Study on the Performance of Resampling Algorithms for Addressing Class Imbalance in Regression Models Husain, G.; Nasef, D.; Jose, R.; Mayer, J.; Bekbolatova, M.; Devine, T.; Toma, M. SMOTE与SMOTEENN：针对回归模型中类别不平衡问题的重采样算法性能研究。Algorithms 2025, 18, 37. [Google Scholar] [CrossRef]

Algorithms

IF0

2025-01-10

PREAI

Gazi Husain; Daniel Nasef; Rejath Jose; Jonathan Mayer; Molly Bekbolatova; Timothy Devine; Milan Toma

Scalable and accurate deep learning with electronic health records 通过电子健康记录进行可扩展且准确的深度学习

NPJ DIGITAL MEDICINE

IF15.1

2018-05-08

1.4K

OAAI

Rajkomar, Alvin; Oren, Eyal; Chen, Kai; Dai, Andrew M.; Hajaj, Nissan; Hardt, Michaela; Liu, Peter J.; Liu, Xiaobing; Marcus, Jake; Sun, Mimi; Sundberg, Patrik; Yee, Hector; Zhang, Kun; Zhang, Yi; Flores, Gerardo; Duggan, Gavin E.; Irvine, Jamie; Quoc Le; Litsch, Kurt; Mossin, Alexander; Tansuwan, Justin; Wang, De; Wexler, James; Wilson, Jimbo; Ludwig, Dana; Volchenboum, Samuel L.; Chou, Katherine; Pearson, Michael; Madabushi, Srinivasan; Shah, Nigam H.; Butte, Atul J.; Howell, Michael D.; Cui, Claire; Corrado, Greg S.; Dean, Jeffrey

Health effects of dietary risks in 195 countries, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017 1990至2017 195国家饮食风险对健康的影响: 全球疾病负担研究2017的系统分析

LANCET

IF88.5

2019-05-01

3.5K

OAAI

Afshin, Ashkan; Sur, Patrick John; Fay, Kairsten A.; Cornaby, Leslie; Ferrara, Giannina; Salama, Joseph S.; Mullany, Erin C.; Abate, Kalkidan Hassen; Abbafati, Cristiana; Abebe, Zegeye; Afarideh, Mohsen; Aggarwal, Anju; Agrawal, Sutapa; Akinyemiju, Tomi; Alahdab, Fares; Bacha, Umar; Bachman, Victoria F.; Badali, Hamid; Badawi, Alaa; Bensenor, Isabela M.; Bernabe, Eduardo; Biryukov, Stan H.; Biadgilign, Sibhatu Kassa K.; Cahill, Leah E.; Carrero, Juan J.; Cercy, Kelly M.; Dandona, Lalit; Dandona, Rakhi; Anh Kim Dang; Degefa, Meaza Girma; Zaki, Maysaa El Sayed; Esteghamati, Alireza; Esteghamati, Sadaf; Fanzo, Jessica; Farinha, Carla Sofia E. Sa; Farvid, Maryam S.; Farzadfar, Farshad; Feigin, Valery L.; Fernandes, Joao C.; Flor, Luisa Sorio; Foigt, Nataliya A.; Forouzanfar, Mohammad H.; Ganji, Morsaleh; Geleijnse, Johanna M.; Gillum, Richard F.; Goulart, Alessandra C.; Grosso, Giuseppe; Guessous, Idris; Hamidi, Samer; Hankey, Graeme J.; Harikrishnan, Sivadasanpillai; Hassen, Hamid Yimam; Hay, Simon I.; Chi Linh Hoang; Horino, Masako; Islami, Farhad; Jackson, Maria D.; James, Spencer L.; Johansson, Lars; Jonas, Jost B.; Kasaeian, Amir; Khader, Yousef Saleh; Khalil, Ibrahim A.; Khang, Young-Ho; Kimokoti, Ruth W.; Kokubo, Yoshihiro; Kumar, G. Anil; Lallukka, Tea; Lopez, Alan D.; Lorkowski, Stefan; Lotufo, Paulo A.; Lozano, Rafael; Malekzadeh, Reza; Marz, Winfried; Meier, Toni; Melaku, Yohannes A.; Mendoza, Walter; Mensink, Gert B. M.; Micha, Renata; Miller, Ted R.; Mirarefin, Mojde; Mohan, Viswanathan; Mokdad, Ali H.; Mozaffarian, Dariush; Nagel, Gabriele; Naghavi, Mohsen; Cuong Tat Nguyen; Nixon, Molly R.; Ong, Kanyin L.; Pereira, David M.; Poustchi, Hossein; Qorbani, Mostafa; Rai, Rajesh Kumar; Razo-Garcia, Christian; Rehm, Colin D.; Rivera, Juan A.; Rodriguez-Ramirez, Sonia; Roshandel, Gholamreza; Roth, Gregory A.; Sanabria, Juan; Sanchez-Pimienta, Tania G.; Sartorius, Benn; Schmidhuber, Josef; Schutte, Aletta Elisabeth; Sepanlou, Sadaf G.; Shin, Min-Jeong; Sorensen, Reed J. D.; Springmann, Marco; Szponar, Lucjan; Thorne-Lyman, Andrew L.; Thrift, Amanda G.; Touvier, Mathilde; Bach Xuan Tran; Tyrovolas, Stefanos; Ukwaja, Kingsley Nnanna; Ullah, Irfan; Uthman, Olalekan A.; Vaezghasemi, Masoud; Vasankari, Tommi Juhani; Vollset, Stein Emil; Theo Vos; Giang Thu Vu; Linh Gia Vu; Weiderpass, Elisabete; Werdecker, Andrea; Wijeratne, Tissa; Willett, Walter C.; Wu, Jason H.; Xu, Gelin; Yonemoto, Naohiro; Yu, Chuanhua; Murray, Christopher J. L.

Comparison of feature importance measures as explanations for classification models 特征重要性度量作为分类模型解释的比较

SN Applied Sciences

IF0

2021-02-03

OAAI

Mirka Saarela; Susanne Jauhiainen

Mortality, morbidity and health in developed societies: a review of data sources

Genus

IF0

2018-01-29

OAAI

Guillaume Wunsch; Catherine Gourbin