Open Access Pub publishes peer-reviewed, free-to-read open-access articles. Showing
articles matching Random Forest — open any to read the full text,
or download the PDF or XML.
Jun 2026 DOI 10.14302/issn.2642-9241.jrd-26-6332
de Melo PhilipCorresponding author
Respiratory diseases remain a major contributor to hospital morbidity and mortality worldwide, particularly among elderly patients and individuals with severe pulmonary compromise. Accurate prediction of respiratory mortality is clinically important for triage, resource allocation, ICU utilization, and early intervention. Traditional statistical models frequently demonstrate limited predictive sensitivity because respiratory mortality is influenced by complex interactions among demographic, diagnostic, physiologic, and severity-related variables. In this study, a machine learning framework was developed to predict in-hospital mortality among patients with respiratory disease using administrative and clinically derived variables, including age, sex, length of stay (LOS), diagnostic descriptions, risk of mortality and severity scores. A Random Forest classifier with balanced class weighting was developed and implemented to address nonlinear relationships and class imbalance within the dataset. Initial modeling demonstrated good overall discrimination performance, with receiver operating characteristic area under the curve (ROC-AUC) values approaching 0.84; however, mortality recall remained limited because deceased patients represented a minority class within the original dataset. To improve mortality detection, a physiologically informed synthetic augmentation strategy was developed. Synthetic clinical variables included oxygen saturation, ICU status, ventilator support, sepsis status, systolic blood pressure, creatinine, and lactate levels. Conditional physiologic consistency rules were incorporated during augmentation to preserve clinically plausible relationships among respiratory failure, hemodynamic instability, and organ dysfunction. The augmented dataset substantially improved model sensitivity and balanced mortality classification performance. Final model evaluation demonstrated strong predictive capability, achieving approximately 97% classification accuracy with balanced precision and recall across mortality classes. Confusion matrix analysis revealed marked reduction in false-negative mortality predictions compared with baseline modeling approaches. Feature importance analysis identified physiologic instability markers, respiratory severity classifications, LOS, and diagnostic respiratory categories as dominant predictors of mortality. These findings suggest that hybrid simulation-augmented machine learning frameworks may provide a valuable strategy for respiratory mortality analytics, particularly in datasets with limited real-world mortality prevalence and incomplete physiologic representation.
Dec 2025 DOI 10.14302/issn.2641-4538.jphi-25-5886
Y. Talbert PatriciaCorresponding author
Objective Diabetes poses significant public health challenges, with many individuals remaining undiagnosed and at risk of complications. This study aimed to evaluate the performance of decision tree ensemble methods for predicting diabetes onset using the Framingham Heart Study Teaching Dataset and to explore sex-specific risk patterns relevant to AI-driven interventions. Methods We analyzed data from 11,627 participants, incorporating demographics, vital signs, smoking status, medication use, and laboratory measures. Random Forest classifiers were developed to predict diabetes incidence at approximately 6-year (Period 2) and 12-year (Period 3) follow-ups. Class imbalance was addressed using undersampling, oversampling, and the Synthetic Minority Over-sampling Technique (SMOTE). Results The models demonstrated robust performance, achieving an Area Under the Curve (AUC) of 0.856 in Period 2, and moderate predictive ability in Period 3 (AUC = 0.732 in males, 0.786 in females). Key predictors included glucose level, BMI, systolic blood pressure, age, and heart rate. Notably, differences emerged in predictive accuracy between men and women, suggesting potential sex-specific vulnerabilities that merit further study. Conclusion Machine learning approaches, particularly Random Forests, show promise for medium- and long-term diabetes risk prediction, supporting early identification and intervention efforts. Future work should focus on hyperparameter tuning and explainability techniques, such as SHapley Additive exPlanations (SHAP) values, to improve model precision, interpretability, and fairness. Equity-focused strategies remain critical to ensure AI-driven tools benefit diverse populations and do not exacerbate existing disparities in diabetes care.
May 2019 DOI 10.14302/issn.2639-3166.jar-19-2780
Masoero GiorgioCorresponding author
Accademia di Agricoltura di Torino, Italy
The inoculation of soil with a bio-fertilizer (BF), with arbuscular mycorrhiza fungi, characterizes a Symbiotic (S) agriculture mode, aimed at promoting the yield and health of crops through modifications in the rhizosphere as well as in the plant phenotype. The main objective of this study was to reduce the incidence of Olive Quick Decline Syndrome (OQDS, involving Xylella fastidiosasubsp.pauca) that afflicts the olive groves in Apulia (Italy). Non-inoculated control (C) plants were compared with Symbiotic (S) plants inoculated with 20 kg ha-1 of Micosat F ®, through a 15 cm deep scarification, in the groves of seven farms covering an area of 27 ha. In addition to a visual observation of 484 plants, to obtain a gradation of the disease severity, some objective rapid type methods were utilized to survey the plants and soil , namely leaf pH, NIR tomoscopy of the leaves, hay-litter-bag probes coupled with NIR spectroscopy and the prediction of soil induced respiration. The fingerprinting of the S and C types of leaves and litter-bags was ascertained by means of the use of a random forest algorithm in the classification matrices. The results on the symptoms appeared variable: they were significantly mitigated in two groves out of six, but they were aggravated in one. All the rapid measurements became essentials in a “holistic” model which was able to explain over 95% of the average mitigation / null / aggravation response to BF inoculation. The holistic model gathers differential and compositional analyses of the leaf (pH, crude protein, water) and of the soil (respiration), but depends mainly on the fingerprinting of the C and S leaves and litter-bags. Two keys were identified for a successful inoculation: a high degree of variability of the soil conditions permitting hospitality for the BF with enhancement of the microbial activity in the S soil (lowering the fingerprint of the control litter-bags) and homogeneity of the leaves (with increases in the fingerprint of the S leaves treated with BF). In short, the inoculation of diseased plants with one BF consortium is far from being the ultimate remedy to mitigate OQDS in all situations. Further studies are needed, at a field level, to clarify the soil hosting capacity and to define the mycorrhizal and / or endophytic * plant * pathogen interactions, even using rapid methods.
Mar 2019 DOI 10.14302/issn.2379-7835.ijn-19-2677
Grootveld MartinCorresponding author
Leicester School of Pharmacy, De Montfort University, The Gateway, Leicester LE1 9BH, United Kingdom
Objectives: Sunflower oil (SFO) is regularly employed for cosmetic, emollient and food frying purposes, the latter representing its foremost use globally. Therefore, full investigations of the molecular composition and quality of SFO products are a major requirement. In this study high-field 1H NMR analysis was employed to explore the molecular composition and authenticities of East African virgin (EAV) SFO products, particularly their acylglycerol fatty acid contents, together with those of selected minor constituents. Results acquired were statistically compared to those obtained on commercially-available, EU-approved refined SFO products via NMR-linked multivariate chemometrics strategies. Methodology: High-field 1H NMR spectra of EAV and refined SFOs (n = 55 and 4 respectively) were acquired at an operating frequency of 400 MHz. Their triacylglycerol fatty acid, triacylglycerol hydrolysis product, and sterol and stanol contents were determined via intelligent frequency bucketing and electronic integration of selected resonances. Univariate analysis-of-variance, and multivariate ROC curve evaluations were conducted to determine the magnitude and statistical significance of analyte concentration differences between these two sample classifications. Further multivariate NMR-linked chemometrics analyses such as principal component, random forest and support vector machine classification analyses were also utilised for this purpose. Key Results: Multicomponent 1H NMR analysis demonstrated that EAV SFOs had significantly higher and lower contents of monounsaturated fatty acids (MUFAs) and polyunsaturated fatty acids (PUFAs), respectively, than those of refined SFOs. Furthermore, significantly higher concentrations of ‘health-friendly’, cholesterol-blocking sterols and stanols were also found in these virgin SFO products. Major Conclusions: 1H NMR analysis provides much valuable molecular information regarding the composition and virginal status of SFOs.The high [MUFA]:[PUFA] content ratio of unrefined EAV SFO products renders them more suitable and safer for commercial or domestic deep-frying episodes than refined SFOs (MUFAs are much more resistant to thermally-induced peroxidation than PUFAs). These products also potentially offer valuable health benefits in view of their high natural sterol and stanol contents.