Article
Oncology
Moniek van Zutphen, Franzel J. B. van Duijnhoven, Evertine Wesselink, Ruud W. M. Schrauwen, Ewout A. Kouwenhoven, Henk K. van Halteren, Johannes H. W. de Wilt, Renate M. Winkels, Dieuwertje E. Kok, Hendriek C. Boshuizen
Summary: The current lifestyle recommendations for cancer survivors are the same as those for the general public, but it's uncertain which lifestyle behaviors are most vital for prognosis. A study on colorectal cancer patients found that consuming sugary drinks was linked to increased recurrence risk, while intake of fruits and vegetables, liquid fat and oil, and animal protein affected all-cause mortality. Further research is needed to confirm these potential associations.
Article
Health Care Sciences & Services
Anthony Devaux, Catherine Helmer, Robin Genuer, Cecile Proust-Lima
Summary: Predicting the individual risk of clinical events using the complete patient history is challenging. This study extends the competing-risk random survival forests method to handle endogenous longitudinal predictors and compute individual event probabilities. The method transforms predictors into fixed features and computes the final event probability using estimators from multiple trees. The study compares the performance of this method to alternative approaches and demonstrates its usefulness in dementia research.
STATISTICAL METHODS IN MEDICAL RESEARCH
(2023)
Article
Computer Science, Theory & Methods
Alberto Archetti, Francesca Ieva, Matteo Matteucci
Summary: Survival analysis is a fundamental tool in medicine for modeling the time until an event of interest occurs. However, in real-world applications, survival data is often incomplete, censored, distributed, and confidential, posing challenges in scalability and privacy preservation. This paper proposes FedSurF++, an extension of the Federated Survival Forest algorithm, which addresses these challenges by constructing random survival forests in heterogeneous federations. The algorithm achieves comparable performance to existing methods while requiring only a single communication round, improving efficiency, robustness, and privacy preservation. Empirical investigations on real-world datasets demonstrate the success of FedSurF++ in healthcare studies, showcasing its potential for improving the scalability and effectiveness of survival analysis while preserving user privacy.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Mathematical & Computational Biology
Valia Baralou, Natasa Kalpourtzi, Giota Touloumi
Summary: This study systematically compares the performance of random forest (RF), random survival forest (RSF), and Cox proportional hazards (Cox-PH) model through a simulation study and real data analysis. The results show that RF generally performs worst, while the performance of RSF and Cox-PH varies depending on different scenarios and assumptions. Models considering survival time show better performance in real data analysis.
BIOMETRICAL JOURNAL
(2023)
Article
Statistics & Probability
Hunyong Cho, Nicholas P. Jewell, Michael R. Kosorok
Summary: ICRF is an iterative tree ensemble method designed for interval censored survival data, addressing splitting bias and updating survival estimates in a self-consistent manner. The method shows high prediction accuracy and uniform consistency in simulations and applications to avalanche and national mortality data.
JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS
(2022)
Article
Automation & Control Systems
Victor Henrique Alves Ribeiro, Roberto Santana, Gilberto Reynoso-Meza
Summary: This paper proposes two novel machine learning algorithms to improve the automatic target recognition system for unmanned aerial vehicles. These models make use of the stochastic procedure of Random Forests and employ the novel Random Vector Functional Link Tree or Extreme Learning Tree for decision split. Experimental results show that the proposed algorithms outperform other state-of-the-art ensemble learning techniques in terms of predictive performance and computational complexity.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Arthur Hoarau, Arnaud Martin, Jean-Christophe Dubois, Yolande Le Gall
Summary: This paper proposes an Evidential Decision Tree and an Evidential Random Forest, which can handle uncertain and imprecise predictions and can predict rich labels. Experimental results showed better performance for the presented methods compared to other evidential models and recent Cautious Random Forests in handling noisy data and effectively uncertainly and imprecisely labeled datasets. The proposed models also offer better robustness and the ability to predict rich labels, which can be used in other approaches such as active learning.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Information Systems
A. Dmitry Devyatkin, G. Oleg Grigoriev
Summary: This paper proposes an algorithm for training kernel decision trees and random forests, which overcomes the limitations of traditional methods in dealing with multidimensional sparse data. Experimental results show that the proposed algorithm outperforms other methods in various tasks, and the selected regularization technique helps reduce overfitting.
Article
Multidisciplinary Sciences
Meizhu Huang, Dapeng Li, Xinyu Cheng, Qing Pei, Zhiyong Xie, Huating Gu, Xuerong Zhang, Zijun Chen, Aixue Liu, Yi Wang, Fangmiao Sun, Yulong Li, Jiayi Zhang, Miao He, Yuan Xie, Fan Zhang, Xiangbing Qi, Congping Shang, Peng Cao
Summary: Appetitive locomotion is important for animals to approach rewards, but the neuronal circuitry controlling it is still unclear. Researchers discovered an excitatory brain circuit from the superior colliculus to the substantia nigra that enhances appetitive locomotion in mice during predatory hunting, depending on the activity of SNc dopamine neurons.
NATURE COMMUNICATIONS
(2021)
Article
Biochemistry & Molecular Biology
Bin-Yan Zhong, Zhi-Ping Yan, Jun-Hui Sun, Lei Zhang, Zhong-Heng Hou, Xiao-Li Zhu, Ling Wen, Cai-Fang Ni
Summary: This study used a machine learning approach based on the RF model to predict one-year disease control for patients with HCC treated with TACE combined with sorafenib, and identified several independent risk factors. The RF model achieved a higher concordance index compared to the logistic regression model, indicating its accuracy in predicting disease control.
FRONTIERS IN MOLECULAR BIOSCIENCES
(2021)
Article
Automation & Control Systems
Jianyuan Sun, Hui Yu, Guoqiang Zhong, Junyu Dong, Shu Zhang, Hongchuan Yu
Summary: In this article, a new random forests algorithm called random Shapley forests (RSFs) is proposed, which uses the Shapley value to evaluate the importance of each feature. The experiments conducted on benchmark and real-world datasets demonstrate that RSFs outperform or are at least comparable to existing consistent RFs, original RFs, and support vector machines.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Statistics & Probability
Yifan Cui, Michael R. Kosorok, Erik Sverdrup, Stefan Wager, Ruoqing Zhu
Summary: Forest-based methods have gained popularity for non-parametric treatment effect estimation. In this study, causal survival forests are introduced to estimate heterogeneous treatment effects in survival and observational settings with right-censored outcomes. The approach relies on orthogonal estimating equations to adjust for both censoring and selection effects under unconfoundedness. Experimental results show that the approach performs well compared to various baselines.
JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY
(2023)
Article
Computer Science, Information Systems
Donghui Yan, Yingjie Wang, Jin Wang, Honggang Wang, Zhenpeng Li
Summary: Random Projection Forests (rpForests) is a method for K-nearest neighbor (KNN) search that combines multiple KNN-sensitive trees constructed through a series of random projections. It has low computational complexity and can be easily parallelized. Experiments show that rpForests achieves remarkable accuracy in terms of fast decaying missing rate of KNNs and discrepancy in the k-th nearest neighbor distances.
IEEE TRANSACTIONS ON BIG DATA
(2021)
Article
Environmental Sciences
Dorota Jozwicki, Puneet Sharma, Ingrid Mann, Ulf-Peter Hoppe
Summary: This paper presents an approach to segment Polar Mesospheric Summer Echoes (PMSE) from datasets obtained from EISCAT VHF radar data. The method involves manual labeling of data into different categories and using random forests algorithm to segment the PMSE. The results show that random forests can effectively segment the PMSE and the weighted-down labels technique improves the performance.
Review
Engineering, Industrial
Anestis Antoniadis, Sophie Lambert-Lacroix, Jean-Michel Poggi
Summary: Understanding physical and engineering problems often requires running complex computational models with a high number of input variables. Global sensitivity analysis (GSA) methods help identify influential inputs, with meta-modeling and random forests providing efficient non-parametric approaches to sensitivity analysis. This allows for dimension reduction and effective quantification of variable importance, even in high-dimensional data sets.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2021)
Article
Surgery
Eileen M. Hsich, Lucy Thuita, Dennis M. McNamara, Joseph G. Rogers, Maryam Valapour, Lee R. Goldberg, Clyde W. Yancy, Eugene H. B. Ackstone, Hemant Ishwaran
AMERICAN JOURNAL OF TRANSPLANTATION
(2019)
Article
Computer Science, Artificial Intelligence
Robert O'Brien, Hemant Ishwaran
PATTERN RECOGNITION
(2019)
Article
Biochemistry & Molecular Biology
Joseph L. Benci, Lexus R. Johnson, Ruth Choa, Yuanming Xu, Jingya Qiu, Zilu Zhou, Bihui Xu, Darwin Ye, Katherine L. Nathanson, Carl H. June, E. John Wherry, Nancy R. Zhang, Hemant Ishwaran, Matthew D. Hellmann, Jedd D. Wolchok, Taku Kambayashi, Andy J. Minn
Article
Oncology
Thomas W. Rice, Min Lu, Hemant Ishwaran, Eugene H. Blackstone
JOURNAL OF THORACIC ONCOLOGY
(2019)
Article
Surgery
Siva Raja, Thomas W. Rice, Sudish C. Murthy, Usman Ahmad, Marie E. Semple, Eugene H. Blackstone, Hemant Ishwaran
Summary: This study demonstrates the importance of lymphadenectomy during esophagectomy for patients with adenocarcinoma of the esophagus and esophagogastric junction undergoing neoadjuvant therapy. Optimal range of nodes resected is associated with maximized survival, with a nonlinear relationship between lymph node resection and lifetime for different ypTNM cancer categories.
Editorial Material
Cardiac & Cardiovascular Systems
Hemant Ishwaran, Robert O'Brien
JOURNAL OF THORACIC AND CARDIOVASCULAR SURGERY
(2021)
Article
Cardiac & Cardiovascular Systems
Eileen M. Hsich, Eugene H. Blackstone, Lucy W. Thuita, Dennis M. McNamara, Joseph G. Rogers, Clyde W. Yancy, Lee R. Goldberg, Maryam Valapour, Gang Xu, Hemant Ishwaran
JACC-HEART FAILURE
(2020)
Letter
Cardiac & Cardiovascular Systems
Hemant Ishwaran, Robert O'Brien
JOURNAL OF THORACIC AND CARDIOVASCULAR SURGERY
(2022)
Editorial Material
Biology
Min Lu, Hemant Ishwaran
Article
Ophthalmology
Robert C. O'Brien, Hemant Ishwaran, Loretta B. Szczotka-Flynn, Jonathan H. Lass
Summary: This study reanalyzes types of intraoperative complications associated with DSAEK graft failure in the Cornea Preservation Time Study using random survival forests, and finds that intraoperative complications are highly predictive of graft failure.
JAMA OPHTHALMOLOGY
(2021)
Article
Computer Science, Artificial Intelligence
Alejandro Mantero, Hemant Ishwaran
Summary: The sidClustering algorithm involves sidification of features followed by prediction using a multivariate random forest, able to handle continuous and categorical variables. This method can effectively identify clusters from various types of variables while retaining the advantages of random forests.
STATISTICAL ANALYSIS AND DATA MINING
(2021)
Article
Cardiac & Cardiovascular Systems
George R. Marzouka, Harold Rivner, Vijay Mehta, Juan Lopez, Igor Vaz, Fei Tang, Hemant Ishwaran, Jeffrey J. Goldberger
Summary: The study showed that the CHA(2)DS(2)-VASc score can effectively stratify the risk of stroke in HF patients and predict the occurrence of stroke in HF patients, regardless of whether they have AF. Patients with higher CHA(2)DS(2)-VASc scores, with or without AF, have increased risk of stroke, thromboembolism, and death.
AMERICAN JOURNAL OF CARDIOLOGY
(2021)
Article
Multidisciplinary Sciences
Min Lu, Hemant Ishwaran
Summary: Various factors influence the dynamics of COVID-19 in the U.S., with economic factors and vaccination efforts playing key roles. A compartmental model framework was developed to analyze the impact of these factors, revealing the importance of cure time and mortality rate. Vaccination data showed suppression of a fourth wave and lower mortality rate for the vaccinated.
Article
Psychiatry
Lloyd D. Balbuena, Marilyn Baetz, Joseph Andrew Sexton, Douglas Harder, Cindy Xin Feng, Kerstina Boctor, Candace LaPointe, Elizabeth Letwiniuk, Arash Shamloo, Hemant Ishwaran, Ann John, Anne Lise Brantsaeter
Summary: This study utilizes machine learning models to identify risk and protective factors for suicide deaths in the general population and clinical samples. It suggests that suicide prevention requires individual actions with governmental incentives, and machine learning can help identify early prevention targets.
Article
Statistics & Probability
Donald K. K. Lee, Ningyuan Chen, Hemant Ishwaran
Summary: This study introduces a generic gradient boosting procedure for survival processes, implemented using regression trees, with regularization methods to prevent overfitting. The research reveals that stepsize restriction serves as a mechanism to prevent risk curvature from derailing convergence.
ANNALS OF STATISTICS
(2021)