Elastic Net Regression Model Research Articles

Abstract 1022: Integration of cerebrospinal fluid methylome and proteome can obviate the need for biopsy in central nervous system lymphoma

Abstract Background: The diagnosis of intra-axial brain tumors requires histopathological examination of tissue obtained by neurosurgery in current clinical practice, which comes with inherent risks. Some patients, particularly those with primary CNS lymphoma (PCNSL), only undergo surgery to obtain a diagnosis and do not derive any therapeutic benefit from surgical resection or debulking. Less-invasive techniques for diagnosis, such as liquid biopsy, therefore provides an opportunity to mitigate surgical risk in these patients. Methods: Patients with histopathology confirmed glioblastoma (GBM, IDH wild type), brain metastases (BM), and PCNSL with accompanying cerebrospinal fluid (CSF) samples were included in our study. Cell-free DNA methylation profiling and shotgun proteomics were obtained for all patients and used to train classifiers to distinguish each tumour entity from others. Specifically, binomial elastic net regression models were built by combining both data modalities using established early and late integration paradigms and performance was compared to classifiers built from individual data types alone. Each model was repeated 100 fold and performance assessed on an untouched testing subset. Results: Our cohort includes 20 patients with GBM, 17 with BM, and 14 with PCNSL, each with matching CSF cell-free DNA methylation and shotgun proteomic profiling. We show that these data can be integrated to fully discriminate PCNSL from its major diagnostic counterparts with a perfect median AUC of 1.00 (95% CI 1-1) and 100% specificity. Integrated "lymphoma vs other" models significantly outperform models trained on methylation or protein data alone, though the same dramatic improvement was not demonstrated in GBM or BM, suggesting synergistic biological information is specific to lymphoma. Conclusions: There is a critical need to diagnose patients with intra-axial tumours, particularly PCNSL, without relying on invasive and costly surgery. We present the most specific and accurate CNS lymphoma classifier to date by integrating the methylome and proteome of CSF. This has the potential for immediate clinical utility, eliminating the need for biopsy in an important subset of these patients. Citation Format: Alex P. Landry, Jeffrey A. Zuccato, Vikas Patil, Mathew Voisin, Justin Z. Wang, Yosef Ellenbogen, Chloe Gui, Andrew Ajisebutu, Farshad Nassiri, Gelareh Zadeh. Integration of cerebrospinal fluid methylome and proteome can obviate the need for biopsy in central nervous system lymphoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1022.

Approximation of Cognitive Performance Using an Elastic Net Regression Model Trained on Gait, Visual, Auditory, Postural, and Olfactory Function Features

Approximation of Cognitive Performance Using an Elastic Net Regression Model Trained on Gait, Visual, Auditory, Postural, and Olfactory Function Features

The Correlation between Metal Mixed Exposure and Lung Function in Different Ages of the Population.

Herein, we explored the overall association between metal mixtures and lung functions in populations of varying ages and the relationship among the associated components. The 2007-2012 National Health and Nutrition Examination Survey data of 4382 American participants was analyzed, and generalized linear, elastic net, quantile g-computation, and Bayesian kernel machine regression models were used to evaluate the relationship between exposure to the metal mixture and lung function at various ages. The results of barium exposure at distinct stages revealed that children and adolescents exhibited greater lung function changes than those in adults and the elderly. Additionally, compared with children and adolescents, cadmium- and arsenic-containing metabolites contributed to nonconductive lung function changes in adults and the elderly exposed to metal mixtures. The results showed that the effects of exposure to metal mixtures on lung function in children and adolescents were predominantly caused by lead and barium. Altogether, children and adolescents were found to be more susceptible to metal-exposure-mediated lung function changes than adults and the elderly.

A new application of Elasticnet regression based near-infrared spectroscopy model: Prediction and analysis of 2,3,5,4′-tetrahydroxy stilbene-2-O-β-D-glucoside and moisture in Polygonum multiflorum

This study constructed a near-infrared spectroscopy (NIRS) quantitative analysis model of 2,3,5,4′-tetrahydroxy stilbene-2-O-β-D-glucoside (TSG) and Polygonum multiflorum's water content based on Elasticnet regression (ER). NIRS data, TSG and the moisture content data sets of 117 samples of P. multiflorum were obtained by using NIR scanning and high-performance liquid chromatography analysis (HPLC) methods. MSE was used to determine the hyperparameters α and λ of Elasticnet. The optimal α values of TSG and water content were 0.83 and 0.01, and the optimal λ values were 0.03 and 1.13, respectively. By preprocessing the original spectral data, the NIRS model was established based on the source data set using ER, compared with the model established by PLS and its optimization algorithm. The results show that the ER model has good performance compared with other methods, with R2c value was 0.9879, RMSEC value was 0.09, R2p value was 0.9784, RMSEP value was 0.13 and RPD value was 6.8. In addition, according to the properties of Elasticnet's variable selection, we speculate that the final coefficients of spectral features may indirectly reflect the spectral features of different components. The results showed better prediction accuracy and superior performance. This is a much more reliable method used to forecast the content of effective components in plants.

Global metabolomics revealed deviations from the metabolic aging clock in colorectal cancer patients.

Background: Markers of aging hold promise in the context of colorectal cancer (CRC) care. Utilizing high-resolution metabolomic profiling, we can unveil distinctive age-related patterns that have the potential to predict early CRC development. Our study aims to unearth a panel of aging markers and delve into the metabolomic alterations associated with aging and CRC. Methods: We assembled a serum cohort comprising 5,649 individuals, consisting of 3,002 healthy volunteers, 715 patients diagnosed with colorectal advanced precancerous lesions (APL), and 1,932 CRC patients, to perform a comprehensive metabolomic analysis. Results: We successfully identified unique age-associated patterns across 42 metabolic pathways. Moreover, we established a metabolic aging clock, comprising 9 key metabolites, using an elastic net regularized regression model that accurately estimates chronological age. Notably, we observed significant chronological disparities among the healthy population, APL patients, and CRC patients. By combining the analysis of circulative carcinoembryonic antigen levels with the categorization of individuals into the "hypo" metabolic aging subgroup, our blood test demonstrates the ability to detect APL and CRC with positive predictive values of 68.4% (64.3%, 72.2%) and 21.4% (17.8%, 25.9%), respectively. Conclusions: This innovative approach utilizing our metabolic aging clock holds significant promise for accurately assessing biological age and enhancing our capacity to detect APL and CRC.

A Comparative Study of Bitcoin’s Price Prediction Using Regression Models

The rising popularity and increasing financial acceptance of cryptocurrency are having a profound impact on global scale. Unlike the current fiat currencies, bitcoins offer a unique possibility to predict their price. Despite the fact that many individuals are investing in cryptocurrencies, little is known about their dynamic properties and predictability, which puts money at risk. The aim of this paper is to evaluate and compare different regression algorithms in order to forecast the price of most popular cryptocurrency – Bitcons. Secondary bitcoin historical data from Kaggle which features an updated daily record of 24 variables over a seven-year period ARE considered. Since the bitcoin data is so volatile, we implemented an effective pre-processing of data in order to have a better prediction result. The different models applied include – Linear Regression, Ridge Regression, LASSO Regression and Elastic Net Regression model. However, elastic net performed better with an RMSE of 0.0228 without showing signs of overfitting.

Proteomic‐based Biological Aging Clock and MRI Markers of Cerebrovascular Disease: Atherosclerosis Risk in Community Study

AbstractBackgroundBiological aging state can be quantified by composite metrics called aging clocks using proteomics. Proteomic aging clocks (PACs) are accurate, easily measured, and are associated with age‐related diseases including Alzheimer’s Disease and Related Dementias. We aim to investigate whether an accelerated biological aging (a discrepancy between chronological age and PAC) is associated with subclinical cerebrovascular structural changes.Method1494 participants from the Atherosclerosis Risk in Communities (ARIC) Study with proteomics and 3T brain MRI data at ARIC visit 5 in 2011‐13 (mean age 76, 59% female, 25% Black) were included. Nearly 5000 plasma proteins were measured using the SomaScan assay. PAC was developed using elastic net regression model and was internally validated. Age acceleration was calculated as residuals after regressing PAC on chronological age (positive value indicates biological age is higher than the person’s chronological age). Linear and logistic regression models were used to assess the associations of age acceleration with white matter hyperintensity volume (in cm3, log2 transformed, median [IQR]: 11[6‐20]) and the presence of: subcortical (n = 281), lacunar (n = 263), and cortical infarcts (n = 150), and microbleeds (n = 355).ResultAccelerated age (median [IQR]:‐0.04[‐1.4,1.3]; correlation with chronological age = 0) was associated with MRI markers of cerebrovascular disease after adjusting for demographic, cardiovascular risk factors, education, and kidney function. Every five years higher age acceleration was associated with larger white matter hyperintensity volume (Difference:0.34[95%CI 0.19,0.49]), higher odds of lacunar (OR:1.61[1.13,2.30]), subcortical (OR:1.63[1.15, 2.32]), and cortical infarcts (OR:1.72[1.11,2.67]). There was no association with presence of microbleeds (OR:1.25[0.91,1.73]) (Figure). Findings were consistent after excluding participants with clinical stroke and there was no difference between APOEe4+ and APOEe4‐ participants.ConclusionHigher accelerated age is cross‐sectionally associated with a greater prevalence of MRI markers of cerebrovascular disease. Understanding this relation has potential to help with risk stratification and personalized prevention and treatment strategies to promote brain health.

An analysis of risk factors for venous thromboembolism in primary versus revision total joint arthroplasty

BackgroundOne of the most serious complications after primary or revision lower extremity total joint arthroplasty (TJA) is venous thromboembolism disease (VTE). Identifying patients at high risk for VTE allows tailoring of prophylactic anticoagulation regimens to those most vulnerable. This study aimed to identify risk factors for VTE in primary and revision lower extremity TJA. MethodsThe Electronic Medical Record was queried from a single academic institution for all patients who underwent a lower extremity TJA between 2007 and 2020. Demographics, comorbid conditions, perioperative characteristics, and postoperative complications were identified. An Elastic Net Multiple Logistic Regression Model was used to assess 49 covariates and predict those associated with a significant risk of VTE. ResultsWe identified 4900 primary and revision total knee arthroplasty (TKA) and total hip arthroplasty (THA) patients. There was no significant difference between primary and revision THA. Primary TKA had a higher rate of VTE than revision TKA. Significant risk factors identified for VTE in THA patients include histories of deep vein thrombosis (DVT), pulmonary embolism (PE), metastatic tumors, hemiplegia, and Hispanic ethnicity. Risk factors for VTE in TKA patients include histories of DVT, PE, metastatic tumors, and postoperative warfarin and heparin use. In all patients, age was a significant predictor of VTE risk. ConclusionOur work identifies many risk factors for VTE following TJA. While the increased rate of VTE in some populations may represent selection bias, it also highlights the incomplete understanding of the etiology and prevention of this complication in the joint arthroplasty population and requires further study.

Patterns of Neural Network Functional Connectivity Associated With Mania/Hypomania and Depression Risk in 3 Independent Young Adult Samples

Mania/hypomania is the pathognomonic feature of bipolar disorder (BD). Established, reliable neural markers denoting mania/hypomania risk to help with early risk detection and diagnosis and guide the targeting of pathophysiologically informed interventions are lacking. To identify patterns of neural responses associated with lifetime mania/hypomania risk, the specificity of such neural responses to mania/hypomania risk vs depression risk, and the extent of replication of findings in 2 independent test samples. This cross-sectional study included 3 independent samples of young adults aged 18 to 30 years without BD or active substance use disorder within the past 3 months who were recruited from the community through advertising. Of 603 approached, 299 were ultimately included and underwent functional magnetic resonance imaging at the University of Pittsburgh, Pittsburgh, Pennsylvania, from July 2014 to May 2023. Activity and functional connectivity to approach-related emotions were examined using a region-of-interest mask supporting emotion processing and emotional regulation. The Mood Spectrum Self-Report assessed lifetime mania/hypomania risk and depression risk. In the discovery sample, elastic net regression models identified neural variables associated with mania/hypomania and depression risk; multivariable regression models identified the extent to which selected variables were significantly associated with each risk measure. Multivariable regression models then determined whether associations in the discovery sample replicated in both test samples. A total of 299 participants were included. The discovery sample included 114 individuals (mean [SD] age, 21.60 [1.91] years; 80 female and 34 male); test sample 1, 103 individuals (mean [SD] age, 21.57 [2.09] years; 30 male and 73 female); and test sample 2, 82 individuals (mean [SD] age, 23.43 [2.86] years; 48 female, 29 male, and 5 nonbinary). Associations between neuroimaging variables and Mood Spectrum Self-Report measures were consistent across all 3 samples. Bilateral amygdala-left amygdala functional connectivity and bilateral ventrolateral prefrontal cortex-right dorsolateral prefrontal cortex functional connectivity were positively associated with mania/hypomania risk: discovery omnibus χ2 = 1671.7 (P < .001); test sample 1 omnibus χ2 = 1790.6 (P < .001); test sample 2 omnibus χ2 = 632.7 (P < .001). Bilateral amygdala-left amygdala functional connectivity and right caudate activity were positively associated and negatively associated with depression risk, respectively: discovery omnibus χ2 = 2566.2 (P < .001); test sample 1 omnibus χ2 = 2935.9 (P < .001); test sample 2 omnibus χ2 = 1004.5 (P < .001). In this study of young adults, greater interamygdala functional connectivity was associated with greater risk of both mania/hypomania and depression. By contrast, greater functional connectivity between ventral attention or salience and central executive networks and greater caudate deactivation were reliably associated with greater risk of mania/hypomania and depression, respectively. These replicated findings indicate promising neural markers distinguishing mania/hypomania-specific risk from depression-specific risk and may provide neural targets to guide and monitor interventions for mania/hypomania and depression in at-risk individuals.

Plasma metabolomics supports non-fasted sampling for metabolic profiling across a spectrum of glucose tolerance in the Nile rat model for type 2 diabetes

Type 2 diabetes is a challenge in modern healthcare, and animal models are necessary to identify underlying mechanisms. The Nile rat (Arvicanthis niloticus) develops diet-induced diabetes rapidly on a conventional rodent chow diet without genetic or chemical manipulation. Unlike common laboratory models, the outbred Nile rat model is diurnal and has a wide range of overt diabetes onset and diabetes progression patterns in both sexes, better mimicking the heterogeneous diabetic phenotype in humans. While fasted blood glucose has historically been used to monitor diabetic progression, postprandial blood glucose is more sensitive to the initial stages of diabetes. However, there is a long-held assumption that ad libitum feeding in rodent models leads to increased variance, thus masking diabetes-related metabolic changes in the plasma. Here we compared repeatability within triplicates of non-fasted or fasted plasma samples and assessed metabolic changes relevant to glucose tolerance in fasted and non-fasted plasma of 8–10-week-old male Nile rats. We used liquid chromatography–mass spectrometry lipidomics and polar metabolomics to measure relative metabolite abundances in the plasma samples. We found that, compared to fasted metabolites, non-fasted plasma metabolites are not only more strongly associated with glucose tolerance on the basis of unsupervised clustering and elastic net regression model, but also have a lower replicate variance. Between the two sampling groups, we detected 66 non-fasted metabolites and 32 fasted metabolites that were associated with glucose tolerance using a combined approach with multivariable elastic net and individual metabolite linear models. Further, to test if metabolite replicate variance is affected by age and sex, we measured non-fasted replicate variance in a cohort of mature 30-week-old male and female Nile rats. Our results support using non-fasted plasma metabolomics to study glucose tolerance in Nile rats across the progression of diabetes.

Neuroanatomical profile of BMI implicates impulsive delay discounting and general cognitive ability.

Obesity is a disorder of excessive adiposity, typically assessed via the anthropometric density measure of BMI. Numerous studies have implicated BMI with differences in brain structure, but with highly inconsistent findings. Machine learning elastic net regression models with cross-validation were conducted to characterize a neuroanatomical morphometry profile associated with BMI in 1100 participants (22% BMI > 30, n = 242) from the Human Connectome Project Young Adult project. Using five-fold cross-validation, the multiregion neuroanatomical profile substantively predicted BMI (R2 = 10.05%), and this was robust in a held-out test set (R2 = 8.87%). In terms of specific regions, the neuroanatomical profile was enriched for nodes in the default mode, executive control, and salience networks. The relationship between the morphometry profile and BMI itself was partially mediated by impulsive delay discounting and general cognitive ability. Taken together, these findings reveal a robust machine learning-derived neuroanatomical profile of BMI, one that comprises nodes in motivational brain networks and suggests the functional links to obesity are via self-regulatory capacity and cognitive function.

Environmental exposure to legacy and emerging per- and polyfluoroalkyl substances and endometriosis in women of childbearing age

Per- and polyfluoroalkyl substances (PFAS) are man-made chemicals known for their adverse effects on humans. Growing concern has risen regarding the reproductive toxicity of PFAS, but whether PFAS affect endometriosis remains to be explored. This hospital-based case-control study included 240 laparoscopic-confirmed endometriosis cases and 334 normal controls in China from 2014 to 2018. Concentrations of thirty-three legacy and emerging PFAS were measured in the plasma samples. Associations between single PFAS and endometriosis were estimated by binary logistic regression. The elastic net regression (ENR) model was used to identify dominant PFAS related to endometriosis. The joint effect of the PFAS mixture on endometriosis was assessed by principal component analysis (PCA), Bayesian kernel machine regression (BKMR), and quantile-based g-computation (q-gcomp). In the single-PFAS model, significant positive associations of PFOA [adjusted odds ratio (95 % CI): 1.22 (1.00, 1.51)], total PFOS [1.19 (1.05, 1.34)] and two branched PFOS isomers [1.16 (1.09, 1.22) for 1m-PFOS; 1.18 (1.04, 1.34) for 6m-PFOS] with increased endometriosis odds were observed. Mixture models showed that the joint effect of PFAS mixture on endometriosis was significant [1.24 (1.05, 1.48)], mainly driven by 1m-PFOS. The PFOS isomers profile suggested a PFOS precursor biotransformation source of 1m-PFOS in our population. Our study suggests that branched isomers of PFOS may be associated with endometriosis.

Development and external validation of multimodal postoperative acute kidney injury risk machine learning models.

To develop and externally validate machine learning models using structured and unstructured electronic health record data to predict postoperative acute kidney injury (AKI) across inpatient settings. Data for adult postoperative admissions to the Loyola University Medical Center (2009-2017) were used for model development and admissions to the University of Wisconsin-Madison (2009-2020) were used for validation. Structured features included demographics, vital signs, laboratory results, and nurse-documented scores. Unstructured text from clinical notes were converted into concept unique identifiers (CUIs) using the clinical Text Analysis and Knowledge Extraction System. The primary outcome was the development of Kidney Disease Improvement Global Outcomes stage 2 AKI within 7 days after leaving the operating room. We derived unimodal extreme gradient boosting machines (XGBoost) and elastic net logistic regression (GLMNET) models using structured-only data and multimodal models combining structured data with CUI features. Model comparison was performed using the receiver operating characteristic curve (AUROC), with Delong's test for statistical differences. The study cohort included 138 389 adult patient admissions (mean [SD] age 58 [16] years; 11 506 [8%] African-American; and 70 826 [51%] female) across the 2 sites. Of those, 2959 (2.1%) developed stage 2 AKI or higher. Across all data types, XGBoost outperformed GLMNET (mean AUROC 0.81 [95% confidence interval (CI), 0.80-0.82] vs 0.78 [95% CI, 0.77-0.79]). The multimodal XGBoost model incorporating CUIs parameterized as term frequency-inverse document frequency (TF-IDF) showed the highest discrimination performance (AUROC 0.82 [95% CI, 0.81-0.83]) over unimodal models (AUROC 0.79 [95% CI, 0.78-0.80]). A multimodality approach with structured data and TF-IDF weighting of CUIs increased model performance over structured data-only models. These findings highlight the predictive power of CUIs when merged with structured data for clinical prediction models, which may improve the detection of postoperative AKI.

Effective Factors under Stock Market Regimes

The Chicago Board Options Exchange (CBOE) Volatility Index (VIX) is a measurement of volatility in the stock market, which is closely associated with the return of the risk premium. This paper categorized the monthly log returns of VIX by Gaussian Mixture Models (GMM) and investigates the driving factors of VIX among the Equity Market Volatility (EMV) trackers under different regimes using the elastic net linear regression model. As a result of categorization, two regimes in log returns of VIX are found. Regime 1 with a lower mean covers most of the months, while regime 2 with a higher mean captures the months of extreme log returns. In months of both regimes, the policy-related factor significantly and independently affects VIX. Another factor that largely affects VIX in regime 1 is the macroeconomy and other factors have little impact on VIX in regime 1. Infectious disease, policy, and government related factors are more important in affecting VIX in regime 2.

Quantifying the Spatiotemporal Variation of NPP of Different Land Cover Types and the Contribution of Its Associated Factors in the Songnen Plain

Net primary productivity (NPP) of vegetation is considered an important indicator for ecological stability and is the main object for analyzing the factors influencing the terrestrial carbon cycle. Recent studies have made clear the changes in the NPP of vegetation and its influencing factors at various scales. However, the variations in NPP based on different land cover types under various natural conditions, along with their driving factors, remain not well understood. In this study, spatial overlay analysis was used to investigate the link among climatic, soil moisture (SM), and topographic parameters and NPP of various land cover types after analyzing the spatial and temporal trends of NPP in the Songnen Plain from 2001 to 2020. Additionally, the contribution of each influence factor to the NPP of different land cover types was calculated using the elastic net regression model. The elastic net regression model eliminates the multicollinearity among the influencing factors while maintaining the model stability, and the R2 of all lands is greater than 0.62, which can effectively quantify the contribution of each influencing factor to NPP. The results show a continuously increasing trend of the overall NPP in the research area over the selected 20 years, and NPP increased most significantly in forest land (FOR). Precipitation (PRE) and NPP showed high correlations in all the different land cover types, while the correlations between NPP and other influencing factors were significantly different. In addition, we found that perennials led to a more significant degree of NPP enhancement, and the effect of topographic conditions on NPP was mainly reflected in differences in moisture conditions due to surface runoff. From the results of the modeling calculations, the cumulative contribution of PRE to NPP ranks first in all land types and is the most vital influencing factor of NPP in the Songnen Plain. SM was an important influence, but the contribution of NPP was greater in land classes with shallow root systems. The results of the study revealed the positive transformation relationship of NPP among land cover types in ecologically fragile areas, which provides a reference for ecological restoration and rationalization of land use structure in zones such as intertwined agricultural and pastoral zones.

Warming and increasing precipitation induced greening on the northern Qinghai-Tibet Plateau

The Qinghai-Tibet Plateau (QTP) has a fragile ecosystem that is sensitive to climate change. Due to the amplifying effect of climate change, the QTP has experienced rapid warming and shifting precipitation in recent decades, profoundly impacting the local ecosystem. However, the specific details of how vegetation responds to these changes were unclear, and the corresponding contributions were poorly quantified. Here, we employed an elastic net regression model to investigate the sensitivity of vegetation to climate factors across multiple time scales and various seasons. The vegetation activity was represented by the enhanced vegetation index (EVI), while climate change was represented by temperature, precipitation, photosynthetically active radiation (PAR), and soil moisture fraction (SMF). During 2000–2020, approximately 50 % of the QTP area showed greening, mainly concentrated in the northern region. Climate change explained approximately 70 % of the variation in vegetation during the growing season, 39 % in spring and 66 % in autumn. Grasslands exhibited the highest sensitive to climate change, with a relative contribution of 83 %, followed by mixed forests (70 %), forests (53 %) and deserts (52 %). Both temperature and precipitation significantly affected vegetation, with relative contributions of 29 % and 22 %, respectively, during the growing season. PAR and SMF had less impact on vegetation, with relative contributions of 8 % and 12 %, respectively. In the greening region, precipitation (26 %) was more important for vegetation growth compared to temperature (25 %). These findings emphasize the importance of precipitation on vegetation on the QTP, providing valuable insights for improving regional ecosystem assessment model and promoting the restoration of fragile ecosystems.

Do Mixtures of Beads with Different Sizes Improve Wet Stirred Media Milling of Drug Suspensions?

The impacts of bead sizes and bead mixtures on breakage kinetics, the number of milling cycles applied to prevent overheating, and power consumption during the nanomilling of drug (griseofulvin) suspensions were investigated from both an experimental and theoretical perspective. Narrowly sized zirconia beads with nominal sizes of 100, 200, and 400 µm and their half-and-half binary mixtures were used at 3000 and 4000 rpm with two bead loadings of 0.35 and 0.50. Particle size evolution was measured during the 3 h milling experiments using laser diffraction. An nth-order breakage model was fitted to the experimental median particle size evolution, and various microhydrodynamic parameters were calculated. In general, the beads and their mixtures with smaller median sizes achieved faster breakage. While the microhydrodynamic model explained the impacts of process parameters, it was limited in describing bead mixtures. For additional test runs performed, the kinetics model augmented with a decision tree model using process parameters outperformed that augmented with an elastic-net regression model using the microhydrodynamic parameters. The evaluation of the process merit scores suggests that the use of bead mixtures did not lead to notable process improvement; 100 µm beads generally outperformed bead mixtures and coarser beads in terms of fast breakage, low power consumption and heat generation, and low intermittent milling cycles.

Determinants of public preferences on low-carbon energy sources: Evidence from the United Kingdom

We empirically derive the determinants of British public preferences for different low-carbon energy sources using machine learning algorithm-based variable selection methods (ridge, lasso, and elastic net regression models). We seek to understand the drivers of support for solar, wind, biomass, and nuclear energy, which are the largest low-carbon energy sources and together account for the majority of UK power generation. Explanatory variables examined include those related to demographics, knowledge, perceptions of climate change, and government policy. We carry out a comparative study by synthesising the results of our independent analyses for each energy source and find that the preferred energy sources vary with respondents’ views on anticipated climate change impacts. Those who believe that potential effects of climate change will be catastrophic tend to prefer renewable energy sources whereas those less concerned about climate change tend to prefer nuclear power. The public also prefers energy sources about which they are more familiar or knowledgeable. Energy transition policies should be adjusted to better consider the factors that drive public acceptance of low-carbon energy sources, including exploring ways to secure policy support that considers energy source-specific characteristics and expanding public awareness of climate change and individual technologies, particularly among certain demographics.

Analysis and Prediction of Relative Humidity Level using Generalized Linear Model

The significance of humidity as a critical climate parameter impacts various sectors, including agriculture, health, and energy, necessitating a comprehensive understanding of its influencing factors. This study investigates the influence of climatic variables such as temperature, rainfall, sunshine duration, wind speed, and wind direction on the humidity levels in DKI Jakarta from 2019 to 2022. The objective is to develop a time-independent predictive model for humidity based on historical climate data. The methodology includes data pre-processing to impute missing values and replace outliers, followed by exploratory data analysis to ascertain variable distribution and inter-relationships. A regression model was initially employed for analysis, with subsequent application of regularization via a generalized linear model to enhance prediction accuracy. Results indicate that temperature, rainfall, sunshine duration, and wind direction significantly impact humidity levels in the investigated period. High inter-variable correlation posed challenges of multicollinearity and overfitting in the initial model. However, the application of regularization, trained with 75% of the historical dataset, mitigated these issues and improved model accuracy. This is evident in the improved Mean Squared Error (MSE) performance metrics of the Elastic-Net Regression Model (12.2), compared to the initial Multiple Regression Model (12.5). These findings hold potential implications for weather forecasting and climate change studies

Data Analysis and Optimal Prediction Model Search based on CTG Data

This paper presents a data analysis and optimal prediction model search of fetal health based on cardiotocography (CTG) data. The objective of this study is to develop an accurate and efficient method to predict fetal health outcomes using CTG data. This paper first analyzes the dataset to identify potential predictors of fetal health and investigate their relationship with fetal distress. The result found a strong relationship between some of the variables in the dataset and fetal health. This paper then uses machine learning algorithms to build and compare several prediction models, including elastic net regression and lasso model. The result of this paper shows that a random forest model performs best in terms of AUC. The model can accurately predict fetal health outcomes with an AUC of 0.925. The finding of this paper suggests that CTG data analysis combined with machine learning algorithms can provide a useful tool for prenatal monitoring and management.