Elastic Net Research Articles

Predicting Methanol Space-Time Yield from CO? Hydrogenation Using Machine Learning: Statistical Evaluation of Penalized Regression Techniques

This study investigates the effectiveness of machine learning techniques, specifically penalized regression models Ridge Regression, Lasso Regression, and Elastic Net Regression in predicting methanol space-time yield (STY) from CO? hydrogenation data. Using a dataset derived from Cu-based catalyst research, the study implemented a comprehensive preprocessing approach, including data cleaning, imputation, outlier removal, and normalization. The models were rigorously evaluated through 10-fold cross-validation and tested on unseen data. Ridge Regression outperformed the other models, achieving the lowest Root Mean Squared Error (RMSE) of 0.7706, Mean Absolute Error (MAE) of 0.5627, and Mean Squared Error (MSE) of 0.5938. In comparison, Lasso and Elastic Net Regression models exhibited higher error metrics. Feature importance analysis revealed that Gas Hourly Space Velocity (GHSV) and Molar Masses of Support significantly influence catalytic activity. These findings suggest that Ridge Regression is a promising tool for accurately predicting methanol production, providing valuable insights for optimizing catalytic processes and advancing sustainable practices in chemical engineering.

Animal health as a function of farmer personality and attitude: using the HEXACO model of personality structure to predict farm-level seropositivity for Fasciola hepatica and Ostertagia ostertagi in dairy cows.

Infections with Fasciola hepatica and Ostertagia ostertagi impinge upon dairy cow health and welfare and represent a major economic factor in livestock industry. Control measures largely rely on the use of anthelminthic drugs. However, reports of anthelmintic resistance necessitate sustainable approaches. Farmer characteristics such as attitude and personality are crucial for the implementation of control strategies and on-farm practices. In the present study, the HEXACO (Honesty-Humility, Emotionality, eXtraversion, Agreeableness, Conscientiousness, Openness to experience) model of personality structure, which conceptualizes human personality, was used to evaluate the relationship of farmer aspects with on-farm bulk tank milk seropositivity for F. hepatica and O. ostertagi. Moreover, information on farm structure, housing, management, and farmers' attitude was collected in a face-to-face interview. Farm-level seropositivity for F. hepatica and O. ostertagi was predicted via elastic net regression. Out of 193 farms housing 8,774 cows in the German Federal State of Bavaria, 47 farms (24.4%) were seropositive for F. hepatica, 77 farms (39.9%) for O. ostertagi, and 42 farms (21.8%) for both endoparasites. The model for F. hepatica seropositivity selected the covariates pasture access, O. ostertagi seropositivity, higher farmer conscientiousness, and organic farming as relevant predictors. Seropositivity for O. ostertagi was predicted by F. hepatica seropositivity, pasture access, organic farming, and farmers being neutral regarding their satisfaction with animal health on their farm. Higher values for the HEXACO factors extraversion and emotionality were inversely associated with O. ostertagi seropositivity. The present work emphasizes the importance of farmer traits in regard to animal health and parasite occurrence. For the effective acceptance and implementation of sustainable control strategies for livestock helminth infections, it is crucial to consider these aspects to holistically address the challenges of managing parasitic diseases. Moreover, tailored communication strategies can be developed incorporating the understanding of individual stockman characteristics and subsequently ensuring encouragement of stakeholders.

Integrating Clinical Variables, Radiomics, and Tumor-derived Cell-Free DNA for Enhanced Prediction of Resectable Esophageal Adenocarcinoma Outcomes

The value of integrating clinical variables, radiomics, and tumor-derived cell-free DNA (cfDNA) for the prediction of survival and response to chemoradiation of patients with resectable esophageal adenocarcinoma is not yet known. Our aim was to investigate if radiomics and cfDNA metrics combined with clinical variables can improve personalized predictions. A cohort of 111 patients with resectable esophageal adenocarcinoma from 2 centers treated with neoadjuvant chemoradiation therapy was used for exploratory retrospective analyses. Models combining the clinical variables of the SOURCE survival model with radiomic features and cfDNA were built using elastic net regression and internally validated using 5-fold cross-validation. Model performance for overall survival (OS) and time to progression (TTP) were evaluated with the C-index and the area under the curve for pathologic complete response. The best-performing baseline models for OS and TTP were based on the combination of SOURCE-cfDNA that reached a C-index of 0.55 and 0.59 compared with 0.44 to 0.45 with SOURCE alone. The addition of restaging positron emission tomography radiomics to SOURCE was the most promising addition for predicting OS (C-index: 0.65) and TTP (C-index: 0.60). Baseline risk stratification was achieved for OS and TTP by combining SOURCE with radiomics or cfDNA, log-rank P < .01. The best-performing combination model for the prediction of pathologic complete response reached an area under the curve of 0.61 compared with 0.47 with SOURCE variables alone. The addition of radiomics and cfDNA can improve the performance of an established survival model. External validity needs to be further assessed in future studies together with the optimization of radiomic pipelines.

Adolescents' Depression Symptoms After Social Distancing and Restrictions: The EpiCoV French Longitudinal Population-Based Cohort

PurposeThe recent worldwide health pandemic and related social distancing and restrictions negatively impacted adolescents' mental health, including depression, underscoring the need for comprehensive understanding beyond immediate impacts. We aimed to identify factors assessed during the first and second lockdowns that were associated with depression symptoms 2 years after restrictions easing. MethodsThis study included 1,946 adolescents, aged 15–18 years, involved in the French Epidémiologie et Conditions de Vie population-based cohort. Depression symptoms were assessed twice, with the nine-item Patient Health Questionnaire, at the beginning of restrictions easing and 1 year later. Logistic regressions were used to estimate associations among sociodemographic, socioeconomics, health, substance use, lockdown conditions, and pandemic-related factors, self-reported during the first and second lockdowns in France, and depression symptoms. Elastic-Net regression was used to study the joint effect of characteristics assessed during both lockdowns. ResultsA total of 472 (25.25%) adolescents reported depression symptoms in the restrictions easing period, 369 (78.18%) were girls and 103 (21.82%) were boys. Factors associated with a higher risk of reporting depression symptoms included being a girl, a smoker, frequently consuming alcohol, having spent lockdowns in a place without outdoor space, poor perceived health, and pre-existing mental disorders. DiscussionFactors associated with depression symptoms after restrictions easing were well known and mostly not related to social distancing and restrictions, suggesting that already vulnerable youth were more likely to be impacted by the stressful pandemic context. Special caution is needed for these adolescents to limit pandemic's impact on long-lasting mental health. These results also underscore the importance of early identification of at-risk adolescents and the development of interventions to enhance their resilience and support in case of future crises.

Using machine and deep learning to predict short-term complications following trigger digit release surgery

BackgroundTrigger finger is a common disorder of the hand characterized by pain and locking of the digits during flexion or extension. In cases refractory to nonoperative management, surgical release of the A1 pulley can be performed. This study evaluates the ability of machine learning (ML) techniques to predict short-term complications following trigger digit release surgery. MethodsA retrospective study was conducted using data for trigger digit release from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) years 2005–2020. Outcomes of interest were 30-day complications and 30-day return to the operating room. Three ML algorithms were evaluated — a Random Forest (RF), Elastic-Net Regression (ENet), and Extreme Gradient Boosted Tree (XGBoost), along with a deep learning Neural Network (NN). Feature importance analysis was performed in the highest performing model for each outcome to identify predictors with the greatest contributions. ResultsWe included a total of 1209 cases of trigger digit release. The best algorithm for predicting wound complications was the RF, with an AUC of 0.64 ± 0.04. The XGBoost algorithm was best performing for medical complications (AUC: 0.70 ± 0.06) and reoperations (AUC: 0.60 ± 0.07). All three models had performance significantly above the AUC benchmark of 0.50 ± 0.00. On our feature importance analysis, age was distinctively the highest contributing predictor of wound complications. ConclusionsMachine learning can be successfully used for risk stratification in surgical patients. Moving forwards, it is imperative for hand surgeons to continue evaluating applications of ML in the field.

One-year employment outcome prediction after traumatic brain injury: A CENTER-TBI study

BackgroundTraumatic brain injury (TBI) can come with long term consequences for functional outcome that can complicate return to work. ObjectivesThis study aims to make accurate patient-specific predictions on one-year return to work after TBI using machine learning algorithms. Within this process, specific research questions were defined: 1 How can we make accurate predictions on employment outcome, and does this require follow-up data beyond hospitalization? 2 Which predictors are required to make accurate predictions? 3 Are predictions accurate enough for use in clinical practice? MethodsThis study used the core CENTER-TBI observational cohort dataset, collected across 18 European countries between 2014 and 2017. Hospitalized patients with sufficient follow-up data were selected for the current analysis (N = 586). Data regarding hospital stay and follow-up until three months post-injury were used to predict return to work after one year. Three distinct algorithms were used to predict employment outcomes: elastic net logistic regression, random forest and gradient boosting. Finally, a reduced model and corresponding ROC-curve was created. ResultsFull models without follow-up achieved an area under the curve (AUC) of about 81 %, which increased up to 88 % with follow-up data. A reduced model with five predictors achieved similar results with an AUC of 90 %. ConclusionThe addition of three-month follow-up data causes a notable increase in model performance. The reduced model - containing Glasgow Outcome Scale Extended, pre-injury job class, pre-injury employment status, length of stay and age - matched the predictive performance of the full models. Accurate predictions on post-TBI vocational outcomes contribute to realistic prognosis and goal setting, targeting the right interventions to the right patients.

Enhancing Gene Expression Predictions Using Deep Learning and Functional Annotations.

Transcriptome-wide association studies (TWAS) aim to uncover genotype-phenotype relationships through a two-stage procedure: predicting gene expression from genotypes using an expression quantitative trait locus (eQTL) data set, then testing the predicted expression for trait associations. Accurate gene expression prediction in stage 1 is crucial, as it directly impacts the power to identify associations in stage 2. Currently, the first stage of such studies is primarily conducted using linear models like elastic net regression, which fail to capture the nonlinear relationships inherent in biological systems. Deep learning methods have the potential to model such nonlinear effects, but have yet to demonstrably outperform linear methods at this task. To address this gap, we propose a new deep learning architecture to predict gene expression from genotypic variation across individuals. Our method utilizes a learnable input scaling layer in conjunction with a convolutional encoder to capture nonlinear effects and higher-order interactions without compromising on interpretability. We further augment this approach to allow for parameter sharing across multiple networks, enabling us to utilize prior information for individual variants in the form of functional annotations. Evaluations on real-world genomic data show that our method consistently outperforms elastic net regression across a large set of heritable genes. Furthermore, our model statistically significantly improved predictive performance by leveraging functional annotations, whereas elastic net regression failed to show equivalent gains when using the same information, suggesting that our method can capture nonlinear functional information beyond the capability of linear models.

Plasma metabolomic signature of healthy lifestyle, structural brain reserve and risk of dementia.

Although the association between healthy lifestyle and dementia risk has been documented, the relationship between a metabolic signature indicative of healthy lifestyle and dementia risk and the mediating role of structural brain impairment remain unknown. We retrieved 136 628 dementia-free participants from UK Biobank. Elastic net regression was used to obtain a metabolic signature that represented lifestyle behaviours. Cox proportional hazard models were fitted to explore the associations of lifestyle-associated metabolic signature with incident dementia. Causal associations between identified metabolites and dementia were investigated using Mendelian randomization. Mediation analysis was also conducted to uncover the potential mechanisms involving 19 imaging-derived phenotypes (brain volume, grey matter volume, white matter volume and regional grey matter volumes). During a follow-up of 12.55 years, 1783 incident cases of all-cause dementia were identified, including 725 cases of Alzheimer's dementia and 418 cases of vascular dementia. We identified 83 metabolites that could represent healthy lifestyle behaviours using elastic net regression. The metabolic signature was associated with a lower dementia risk, and for each standard deviation increment in metabolic signature, the hazard ratio was 0.89 [95% confidence interval (CI): 0.85, 0.93] for all-cause dementia, 0.95 (95% CI: 0.88, 1.03) for Alzheimer's dementia and 0.84 (95% CI: 0.77, 0.91) for vascular dementia. Mendelian randomization revealed potential causal associations between the identified metabolites and risk of dementia. In addition, the specific structural brain reserve, including the hippocampus, grey matter in the hippocampus, parahippocampal gyrus and middle temporal gyrus, were detected to mediate the effects of metabolic signature on dementia risk (mediated proportion ranging from 6.21% to 11.98%). The metabolic signature associated with a healthy lifestyle is inversely associated with dementia risk, and greater structural brain reserve plays an important role in mediating this relationship. These findings have significant implications for understanding the intricate connections between lifestyle, metabolism and brain health.

The Effects of Adverse Life Events on Brain Development in the ABCD Study®: A Propensity-weighted Analysis.

Longitudinal studies of the effects of adversity on human brain development are complicated by the association of stressful events with confounding variables. To counter this bias, we apply a propensity-weighted analysis of the first two years of The Adolescent Brain Cognitive DevelopmentSM (ABCD) Study® data, employing a machine learning analysis weighted by individuals' propensity to experience adversity. Data included 338 resting-state functional connections from 7190 youth (46% female), divided into a training group (80%) and an independent testing group (20%). Propensity scores were computed using 390 variables to balance across two-year adverse life event exposures. Using elastic net regularization with and without inverse propensity weighting, we developed linear models in which changes in functional connectivity of brain connections during the two-year period served as predictors of the number of adverse events experienced during that same period. Haufe's method was applied to forward-transform the backward prediction models. We also tested whether brain changes associated with adverse events correlated with concomitant changes in internalizing or externalizing behaviors or to academic achievement. In the propensity-weighted analysis, brain development significantly predicted the number of adverse events experienced during that period in both the training group (ρ=0.14, p<0.001) and the independent testing group (ρ=0.10, p<0.001). The predictor indicated a general pattern of decreased functional connectivity between large-scale networks and subcortical brain regions, particularly for cingulo-opercular and sensorimotor networks. These network-to-subcortical functional connectivity decreases inversely associated with the development of internalizing symptoms, suggesting adverse events promoted adaptive brain changes that may buffer against stress-related psychopathology. However, these same functional connections were also associated with poorer grades at the two-year follow-up. Although cortical-subcortical brain developmental responses to adversity potentially shield against stress-induced mood and anxiety disorders, they may be detrimental to other domains such as academic success.

Proteomic Assessment of the Risk of Secondary Cardiovascular Events among Individuals with CKD.

Cardiovascular risk models have been developed primarily for incident events. Well-performing models are lacking to predict secondary cardiovascular events among people with a history of coronary heart disease, stroke, or heart failure who also have chronic kidney disease (CKD). We sought to develop a proteomics-based risk score for cardiovascular events in individuals with CKD and a history of cardiovascular disease. We measured 4638 plasma proteins among 1067 participants from the Chronic Renal Insufficiency Cohort (CRIC) and 536 individuals from the Atherosclerosis Risk in Communities Cohort (ARIC). All had non-dialysis-dependent CKD and coronary heart disease, heart failure, or stroke at study baseline. A proteomic risk model for secondary cardiovascular events was derived by elastic net regression in CRIC, validated in ARIC, and compared to clinical models. Biologic mechanisms of secondary events were characterized through proteomic pathway analysis. A 16-protein risk model was superior to the Framingham risk score for secondary events, including a modified score that included estimated glomerular filtration rate (eGFR). In CRIC, the annualized area under the receiver operating characteristic (AUC) within 1 to 5 years ranged between 0.77 and 0.80 for the protein model and 0.57 and 0.72 for the clinical models. These findings were replicated in the ARIC validation cohort. Biologic pathway analysis identified pathways and proteins for cardiac remodeling and fibrosis, vascular disease, and thrombosis. The proteomic risk model for secondary cardiovascular events outperformed clinical models based on traditional risk factors and eGFR.

Zero-Inflated Binary Classification Model with Elastic Net Regularization

Zero-Inflated Binary Classification Model with Elastic Net Regularization

Comparative analysis of some linear predictive models in the presence of multicollinearity

This paper investigates the comparative performance of some linear predictive models in the presence of multicollinearity. By examining the efficacy of Ordinary Least Squares (OLS), Ridge Regression, Lasso Regression, and Elastic Net Regression, this study aimed to figure out the best method for building robust and interpretable models under such conditions. The research explores how these models address multicollinearity, focusing on coefficient stability, prediction accuracy, and variable selection. Through a rigorous analysis of simulated and real-world datasets, the study shows the strengths and weaknesses of each model, providing valuable insights for researchers and practitioners looking to mitigate the challenges posed by multicollinearity in selecting the most proper method for regression modeling. This will lead to the creation of a model with increased interpretability of the relationships between variables, less variance, and more dependable coefficient estimations.

Elastic Net Principal Component Regression With an Application

Elastic Net Principal Component Regression With an Application

Retrospective cohort study of self-reported outcomes after unilateral dacryocystorhinostomy in patients with bilateral epiphora and suspected nasolacrimal obstruction

ABSTRACT Purpose This study aimed to quantify the proportion of patients with bilateral epiphora and suspected nasolacrimal obstruction who self-reported improved epiphora bilaterally following unilateral dacryocystorhinostomy and investigate predictive factors. Methods A retrospective cohort study of patients from an Australian lacrimal clinic was performed. Eligible patients were at least 18 years of age, complained of bilateral epiphora, underwent unilateral dacryocystorhinostomy between 2012 and 2022 and followed-up. Outcome groups were created by grouping self-reported epiphora improvement levels for each eye. Two-sided confidence intervals were calculated for the proportion of participants in each outcome group. Ordinal elastic net regression identified clinical features associated with outcome groups and estimated their effect sizes. Results Of 243 patients, the median post-operative follow-up duration was 3 months. Fifty-eight percent (95% CI 52–64%, n = 141) reported significant improvement (≥50% ipsilateral and any contralateral improvement), 5% (95% CI 3–9%, n = 13) reported small improvement (<50% ipsilateral and any contralateral improvement) and 26% (95% CI 20–31%, n = 62) reported ipsilateral improvement only. Eleven percent had functional or anatomical failure. Older age (OR 1.01), contralateral nasolacrimal duct narrowing (OR 1.37), contralateral nasolacrimal duct obstruction (OR 0.93) and longer follow-up time (OR 0.85) predicted outcome groups. Conclusions After unilateral DCR, many patients with bilateral epiphora and suspected nasolacrimal obstruction report an improvement of their symptoms bilaterally in the early postoperative period. Contralateral nasolacrimal system diagnosis, patient age and follow-up duration were associated with outcomes. Further study using validated epiphora assessment tools may quantify this observation, and we suggest a potential underlying mechanism of contralateral effects after unilateral treatment.

Simulating multi-scale optimization and variable selection in species distribution modeling

Species distribution modeling (SDM) is a fundamental tool in theoretical and applied ecology. However, relatively little is known about the performance of different approaches for scale optimization, model selection, and algorithmic prediction in the context of nonlinear, multiscale and interactive relationships between environmental variables and species occurrence. Modelers often struggle to optimize a tradeoff between ecological relevance, model robustness, complexity, and overfitting. In this paper, we investigated several methods designed to optimize spatial scale and variable selection in SDMs, in each case evaluating model fitness, parsimony and predictive performance. We used a simulation approach to produce a large pool of alternative underlying habitat relationships that reflect a broad range of realistic habitat associations. We also compared several different modeling algorithms, including logistic regression with a generalized linear model (GLM), Lasso and Elastic-Net Regularized GLMs (GLMNet), and random forest (RF), as well as alternative variable and scale selection methods. We found that GLM methods employing all-subsets dredge routines for variable selection were consistently the best predictors based on all criteria of our model performance assessment and across all attributes of the simulated underlying relationship, including nonlinearity and interaction. We had expected machine learning approaches, such as random forest, to perform better in these more complex forms of species-environment relationships. GLM using dredge variable selection was also the method that included the fewest spurious covariates and included the most correct predictors as a proportion of all predictors. We found that univariate scaling was the most robust method of variable and scale selection, along with Minimal Redundancy Maximal Relevancy (MRMR) which performed equivalently. The simulation experiment presented here provides a robust assessment of simulated multi-species distribution model performance, complexity and fidelity. By simulating a large range of potential habitat relationships with varying spatial scale, effect sizes, linearity, and interactions, we comprehensively evaluated model performance across gradients of complexity of the underlying relationships and violations of classical statistical assumptions. This study provides a valuable assessment and a broader example of the power and utility of controlled simulation experiments in habitat relationships and other ecological spatial predictive modeling.

Latent profiles and predictors of barriers to care in Swiss children and adolescents with rare diseases.

Children and adolescents with rare diseases face significant barriers when accessing healthcare. We aimed to assess and predict these barriers and investigate associations with health-related quality of life (HRQoL). We conducted a cross-sectional survey of Swiss parents (N = 189) of children with rare diseases including the Barriers to Care Questionnaire (BCQ), containing six barriers and the Pediatric Quality of Life Inventory (PedsQL). Latent profile analysis (LPA) was used to uncover distinct classes, which were compared using chi-square tests and Mann-Whitney U tests. Relevant medical and sociodemographic class predictors were identified using Elastic Net regression, followed by regression analysis to investigate their role in predicting barriers to care and examine the effects of these classes on HRQoL. Two distinct groups were identified, a higher barriers class (59%) and a lower barriers class (41%). In the higher barriers class, participants showed elevated scores across all subscales and specifically on pragmatics and expectations. More barriers to care were linked to a nonstable disease course (OR = 2.27, p = .002) and a diagnosis after the age of 3months (OR = 2.17, p = .006). Individuals in the higher barriers class exhibited more psychological comorbidities (p = .044), congenital malformations/deformations/chromosomal abnormalities (p=.042), and medical misdiagnoses (p = .006). Children in the higher barriers class had significantly lower PedsQL scores compared to the lower barriers class (p<.05). This study highlights the need for comprehensive assessment of barriers to pediatric care in rare diseases, offering potential entry points for targeted interventions.

Development, characterization, and replication of proteomic aging clocks: Analysis of 2 population-based cohorts.

Biological age may be estimated by proteomic aging clocks (PACs). Previous published PACs were constructed either in smaller studies or mainly in white individuals, and they used proteomic measures from only one-time point. In this study, we created de novo PACs and compared their performance to published PACs at 2 different time points in the Atherosclerosis Risk in Communities (ARIC) study of white and black participants (around 75% white and 25% black). A total of 4,712 plasma proteins were measured using SomaScan in blood samples collected in 1990 to 1992 from 11,761 midlife participants (aged 46 to 70 years) and in 2011 to 2013 from 5,183 late-life participants (aged 66 to 90 years). The de novo ARIC PACs were constructed by training them against chronological age using elastic net regression in two-thirds of healthy participants in midlife and late life and validated in the remaining one-third of healthy participants at the corresponding time point. We also computed 3 published PACs. We estimated age acceleration for each PAC as residuals after regressing each PAC on chronological age. We also calculated the change in age acceleration from midlife to late life. We examined the associations of age acceleration and change in age acceleration with mortality through 2019 from all-cause, cardiovascular disease (CVD), cancer, and lower respiratory disease (LRD) using Cox proportional hazards regression in participants (irrespective of health) after excluding the training set. The model was adjusted for chronological age, smoking, body mass index (BMI), and other confounders. We externally validated the midlife PAC using the Multi-Ethnic Study of Atherosclerosis (MESA) Exam 1 data. The ARIC PACs had a slightly stronger correlation with chronological age than published PACs in healthy participants at each time point. Associations with mortality were similar for the ARIC PACs and published PACs. For late-life and midlife age acceleration for the ARIC PACs, respectively, hazard ratios (HRs) per 1 standard deviation were 1.65 and 1.38 (both p < 0.001) for all-cause mortality, 1.37 and 1.20 (both p < 0.001) for CVD mortality, 1.21 (p = 0.028) and 1.04 (p = 0.280) for cancer mortality, and 1.68 and 1.36 (both p < 0.001) for LRD mortality. For the change in age acceleration, HRs for all-cause, CVD, and LRD mortality were comparable to the HRs for late-life age acceleration. The association between the change in age acceleration and cancer mortality was not significant. The external validation of the midlife PAC in MESA showed significant associations with mortality, as observed for midlife participants in ARIC. The main limitation is that our PACs were constructed in midlife and late-life participants. It is unknown whether these PACs could be applied to young individuals. In this longitudinal study, we found that the ARIC PACs and published PACs were similarly associated with an increased risk of mortality. These findings suggested that PACs show promise as biomarkers of biological age. PACs may be serve as tools to predict mortality and evaluate the effect of anti-aging lifestyle and therapeutic interventions.

Optimized Machine Learning Models for Predicting Core Body Temperature in Dairy Cows: Enhancing Accuracy and Interpretability for Practical Livestock Management.

Heat stress poses a significant challenge to livestock farming, particularly affecting the health and productivity of high-yield dairy cows. This study develops a machine learning framework aimed at predicting the core body temperature (CBT) of dairy cows to enable more effective heat stress management and enhance animal welfare. The dataset includes 3005 records of physiological data from real-world production environments, encompassing environmental parameters, individual animal characteristics, and infrared temperature measurements. Employed machine learning algorithms include elastic net (EN), artificial neural networks (ANN), random forests (RF), extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), and CatBoost, alongside several optimization algorithms such as Bayesian optimization (BO) and grey wolf optimizer (GWO) to refine model performance through hyperparameter tuning. Comparative analysis of various feature sets reveals that the feature set incorporating the average infrared temperature of the trunk (IRTave_TK) excels in CBT prediction, achieving a coefficient of determination (R2) value of 0.516, mean absolute error (MAE) of 0.239 °C, and root mean square error (RMSE) of 0.302 °C. Further analysis shows that the GWO-XGBoost model surpasses others in predictive accuracy with an R2 value of 0.540, RMSE as low as 0.294 °C, and MAE of just 0.232 °C, and leads in computational efficiency with an optimization time of merely 2.41 s-approximately 4500 times faster than the highest accuracy model. Through SHAP (SHapley Additive exPlanations) analysis, IRTave_TK, time zone (TZ), days in lactation (DOL), and body posture (BP) are identified as the four most critical factors in predicting CBT, and the interaction effects of IRTave_TK with other features such as body posture and time periods are unveiled. This study provides technological support for livestock management, facilitating the development and optimization of predictive models to implement timely and effective interventions, thereby maintaining the health and productivity of dairy cows.

BayesAge 2.0: A Maximum Likelihood Algorithm to Predict Transcriptomic Age.

Aging is a complex biological process influenced by various factors, including genetic and environmental influences. In this study, we present BayesAge 2.0, an improved version of our maximum likelihood algorithm designed for predicting transcriptomic age (tAge) from RNA-seq data. Building on the original BayesAge framework, which was developed for epigenetic age prediction, BayesAge 2.0 integrates a Poisson distribution to model count-based gene expression data and employs LOWESS smoothing to capture non-linear gene-age relationships. BayesAge 2.0 provides significant improvements over traditional linear models, such as Elastic Net regression. Specifically, it addresses issues of age bias in predictions, with minimal age-associated bias observed in residuals. Its computational efficiency further distinguishes it from traditional models, as reference construction and cross-validation are completed more quickly compared to Elastic Net regression, which requires extensive hyperparameter tuning. Overall, BayesAge 2.0 represents a notable advance in transcriptomic age prediction, offering a robust, accurate, and efficient tool for aging research and biomarker development.

Robust multi-target regression with improved stochastic configuration networks and its applications

To improve the accuracy and robustness of stochastic configuration networks (SCNs) for resolving multi-target regression tasks, this paper proposes a robust modeling approach based on improved stochastic configuration networks. A parallel implementation of SCN models is designed to incrementally generate the hidden nodes, which enhances the diversity of hidden layer mapping through information superposition and spanning connection. We employ an elastic net regularization model to sparsely constrain the model parameters to characterize the correlation among multiple targets. Then, the mixture Laplace distributions are used as the prior distribution of each target modeling error, and the output weights of the SCN model are re-evaluated by maximizing a posteriori estimation to enhance model’s robustness with respect to some uncertainties presented in training samples. The modelling performance of the proposed solution is tested on six standard datasets and the historical data of a municipal solid waste incineration process. The experimental results show that the proposed modeling technique has advantages in terms of both the prediction accuracy and the robustness.