IISE Transactions最新文献

The increasing availability of health data from resources such as large biobanks, electronic healthcare records, medical tests, and wearable sensors, has set the stage for the development of novel machine learning (ML) models for multi-modal mixed-type data to capture the complexity of human health and disease. Clustering is a type of ML model that aims to identify homogenous subgroups from heterogeneous data, providing a data-driven solution to targeted, subgroup-specific study and intervention. While such data contain diverse and complementary information to facilitate decision making and improve population health, clustering of high-dimensional multi-modal mixed-type data poses major challenges to existing ML and statistical models. We propose a novel Multi-modal Mixed-type Structural Equation Model (M2-SEM) with structured sparsity to cluster heterogeneous health data for precise subgroup discovery. To accommodate a mix of continuous and categorical data modalities, we developed a novel Gauss-Hermite-enabled Expectation-Majorization-Minimization (GH-EMM) algorithm that integrates the GH quadrature and the Majorization Maximization (MM) algorithm within the Expectation Maximization (EM) framework for efficient model estimation. The proposed M2-SEM and GH-EMM are first tested in extensive simulation studies in comparison with benchmarks, and then applied to identify subgroups of individuals with low- and high-risk of developing adverse cardiometabolic (CM) outcomes based on a full spectrum of CM risk factors such as poor nutrition and mental health, physical inactivity, and sleep deprivation. These findings shed light on the promise of using multi-modal mixed-type health data for early identification and targeted intervention of at-risk individuals for health promotion at the population level.

Combining multiple predictors obtained from distributed data sources to an accurate meta-learner is promising to achieve enhanced performance in lots of prediction problems. As the accuracy of each predictor is usually unknown, integrating the predictors to achieve better performance is challenging. Conventional ensemble learning methods assess the accuracy of predictors based on extensive labeled data. In practical applications, however, the acquisition of such labeled data can prove to be an arduous task. Furthermore, the predictors under consideration may exhibit high degrees of correlation, particularly when similar data sources or machine learning algorithms were employed during their model training. In response to these challenges, this paper introduces a novel structured unsupervised ensemble learning model (SUEL) to exploit the dependency between a set of predictors with continuous predictive scores, rank the predictors without labeled data and combine them to an ensembled score with weights. Two novel correlation-based decomposition algorithms are further proposed to estimate the SUEL model, constrained quadratic optimization (SUEL.CQO) and matrix-factorization-based (SUEL.MF) approaches. The efficacy of the proposed methods is rigorously assessed through both simulation studies and real-world application of risk genes discovery. The results compellingly demonstrate that the proposed methods can efficiently integrate the dependent predictors to an ensemble model without the need of ground truth data.