Lasso Regression Methods Research Articles

Potential diagnostic markers and therapeutic targets for non-alcoholic fatty liver disease and ulcerative colitis based on bioinformatics analysis and machine learning.

Non-alcoholic fatty liver disease (NAFLD) and ulcerative colitis (UC) are two common health issues that have gained significant global attention. Previous studies have suggested a possible connection between NAFLD and UC, but the underlying pathophysiology remains unclear. This study investigates common genes, underlying pathogenesis mechanisms, identification of diagnostic markers applicable to both conditions, and exploration of potential therapeutic targets shared by NAFLD and UC. We obtained datasets for NAFLD and UC from the GEO database. The DEGs in the GSE89632 dataset of the NAFLD and GSE87466 of the UC dataset were analyzed. WGCNA, a powerful tool for identifying modules of highly correlated genes, was employed for both datasets. The DEGs of NAFLD and UC and the modular genes were then intersected to obtain shared genes. Functional enrichment analysis was conducted on these shared genes. Next, we utilize the STRING database to establish a PPI network. To enhance visualization, we employ Cytoscape software. Subsequently, the Cytohubba algorithm within Cytoscape was used to identify central genes. Diagnostic biomarkers were initially screened using LASSO regression and SVM methods. The diagnostic value of ROC curve analysis was assessed to detect diagnostic genes in both training and validation sets for NAFLD and UC. A nomogram was also developed to evaluate diagnostic efficacy. Additionally, we used the CIBERSORT algorithm to explore immune infiltration patterns in both NAFLD and UC samples. Finally, we investigated the correlation between hub gene expression, diagnostic gene expression, and immune infiltration levels. We identified 34 shared genes that were found to be associated with both NAFLD and UC. These genes were subjected to enrichment analysis, which revealed significant enrichment in several pathways, including the IL-17 signaling pathway, Rheumatoid arthritis, and Chagas disease. One optimal candidate gene was selected through LASSO regression and SVM: CCL2. The ROC curve confirmed the presence of CCL2 in both the NAFLD and UC training sets and other validation sets. This finding was further validated using a nomogram in the validation set. Additionally, the expression levels of CCL2 for NAFLD and UC showed a significant correlation with immune cell infiltration. This study identified a gene (CCL2) as a biomarker for NAFLD and UC, which may actively participate in the progression of NAFLD and UC. This discovery holds significant implications for understanding the progression of these diseases and potentially developing more effective diagnostic and treatment strategies.

Rapid Prediction and Parameter Evaluation of Process-Induced Deformation in L-Shape Structures Based on Feature Selection and Artificial Neural Networks

The process-induced deformation (PID) during the manufacturing of thermosetting composite materials can significantly compromise manufacturing precision. This paper introduces an innovative method that combines a finite element analysis (FEA), feature classification algorithms, and an Artificial Neural Network (ANN) framework to rapidly predict the PID of a typical L-shaped structure. Initially, a comprehensive range of parameters that influence PID are compiled in this research, followed by the generation of a dataset through FEA considering viscoelastic constitutive models, validated by experimental results. Influential parameters are classified using Random Forest and LASSO regression methods, with each parameter rated according to its impact on PID, delineating their varying degrees of importance. Subsequently, through a hyperparameter analysis, an ANN framework is developed to rapidly predict the PID, while also refining the assessment of the parameters’ significance. This innovative approach achieves a computational time reduction of 98% with less than a 5% loss in accuracy, and highlights that under limited computational conditions, considering only a subset or all of the parameters—the peak temperature, corner angle, coefficient of chemical shrinkage, coefficient of thermal expansion, curing pressure, and E1—minimizes accuracy loss. The study demonstrates that machine learning algorithms can effectively address the challenge of predicting composite material PID, providing valuable insights for practical manufacturing applications.

Determinants for predicting zero-leverage decisions: A machine learning approach

The zero-leverage (ZL) phenomenon is widespread and receives much attention; however, its determinants remain unknown. Using random forest and LASSO regression methods, this study investigates the factors contributing to the ZL phenomenon. We are the first to show that the determinants of overall leverage cannot be directly applied to ZL companies. Findings reveal that cash holdings, tangible assets, industry leverage-level, and firm size are key determinants of ZL. Notably, compared with related studies, ZL shares only some of the determinants of overall leverage, despite being its component. Cash holdings are a determinant unique to ZL companies and the most important among all variables. Using machine learning methods, we identified determinants that are important and reliable, filling a critical gap in relevant research. Moreover, we demonstrate how sample imbalance affects the model’s ability to correctly identify ZL companies and propose a solution to this problem.

Development and validation of a nomogram for sleep disorders among stroke patients

BackgroundPrecisely identifying high-risk sleep disorder patients and implementing suitable measures are important for decreasing the incidence of sleep disorders. In this study, a nomogram method was adopted to construct a tool to predict sleep disorders in stroke based on four factors: individual characteristics, treatment-related factors, psychological factors, and family-related factors. MethodsA total of 450 stroke patients were continuously diagnosed at the Affiliated Hospital of Nantong University, and the data on participants were randomly distributed into a training set (n = 315) and a validation set (n = 135). Within the training set, using LASSO regression and random forest methods, five optimal predictors of sleep disorders were identified. Five optimal predictors were used to develop a model. The calibration, discrimination, generalization, and clinical applicability of the model were evaluated using calibration curves, receiver operating characteristic (ROC) curves, internal validation, and decision curve analysis (DCA). ResultsWe found that the place of residence, average daily infusion time, the Hospital Anxiety and Depression Scale (HADS), the Type D Personality Scale-14 (DS14), and the Fatigue Severity Scale (FSS) were crucial factors associated with sleep disorders. The validation data showed an area under the curve (AUC) of 0.903 compared to 0.899 in the training set. There was an approach to the diagonal in the calibration curve of this model, and the results of DCA noted that it is clinically beneficial across a range of thresholds from 5 % to 99 %. ConclusionA model was developed to predict sleep disorders among stroke patients to help hospital staff evaluate the risk among patients and screen high-risk patients.

Strain Gauge Location Optimization for Operational Load Monitoring of an Aircraft Wing Using an Improved Correlation Measure

Operational loads of an aircraft are the prerequisite for assessing its safety or fatigue life. Traditionally, numerous strain gauge sensors are installed to monitor the operational loads, which inevitably increase the weight and system complexity of the aircraft. Therefore, in order to decrease the maintenance costs and data redundancy, the number and location of strain sensors should be optimized for accurate and reliable operational load monitoring. In this paper, a novel two-stage strain gauge location optimization method is proposed to reduce the number of strain gauges while maintaining the operational load monitoring accuracy, which is validated by a numerical case study of an aircraft wing. In the first stage, the traditional Pearson correlation measure is harnessed to initially eliminate numerous correlated strain gauge monitoring points, reducing 996 original strain gauge measurement points to 13 for the aircraft wing box. In the second stage, an improved correlation measure method is proposed to further reduce the 13 strain gauge points to 2, which can evaluate the correlation degree of several variables and simultaneously determine the optimal strain monitoring locations for the two load actuators in this study. The relative errors between the predicted loads and the actual loads for both load actuators are less than 4% when only two optimized monitoring points are adopted. In addition, a comparison study with LASSO regression and principal component regression methods is conducted. The results demonstrate that the proposed method has the characteristics of less monitoring points and higher load prediction precision.

Unravelling the effects of straw return on rice production in central China: Evidence for future policy‐making

AbstractThe practice of straw return represents an effective and sustainable approach to recycling and reusing agricultural residues, thereby significantly contributing to the circular economy within the agri‐food sector. Although this practice holds significant promise for enhancing soil quality and mitigating pollution resulting from straw burning, the realization of these substantial benefits depends on the extent to which farmers are willing to embrace this practice, a decision closely tied to the anticipated economic benefits it offers. The present study aimed to explore the effect of straw return on rice production in terms of rice yield and net return using a representative household survey of 1752 rice farmers in central China. The results showed that there exist threshold effects of the years of straw return on both rice yield and net income, which decline as years of straw return increase below the thresholds, and then increase linearly above the thresholds. Overall, after implementing the straw return practice, it took an average of about 3.5 years to observe a positive effect on rice yields and about 7.5 years to see an increase in net income. The estimates were robust to threshold regression, Lasso regression and random forest regression methods. Nonetheless, the provision of essential technical guidance to farmers via agricultural cooperatives could contribute to expediting the realization of the benefits associated with implementing straw return practices. The results imply that the current incentive mechanism aimed at diffusing straw return practice should be fine‐tuned by considering the time required for straw return to generate economic benefits in specific contexts, ensuring better compensation for initial losses. Interventions that shorten the time needed to realize these benefits, such as providing additional technical training for farmers, should be implemented. Furthermore, the uncertainty regarding the time frame for benefit realization should be communicated clearly and effectively to farmers.

Comprehensive evaluation of immunological attributes and immunotherapy responses of positive T cell function regulators in colorectal cancer

BackgroundPositive regulators of T-cell function (PTFRs), integral to T-cell proliferation and activation, have been identified as potential prognostic markers in colorectal cancer (CRC). Despite this, their role within the tumor microenvironment (TME) and their response to immunotherapy are not yet fully understood.MethodsThis study delved into PTFR-related CRC subtypes by analyzing four independent transcriptome datasets, emphasizing the most significant prognostic PTFRs. We identified differentially expressed genes (DEGs) between two subtypes and developed a PTFR risk model using LASSO and Cox regression methods. The model’s associations with survival time, clinical features, TME characteristics, tumor mutation profiles, microsatellite instability (MSI), cancer stem cell (CSC) index, and responses to chemotherapy, targeted therapy, and immunotherapy were subsequently explored.ResultsThe PTFR risk model demonstrated a strong predictive capacity for CRC. It facilitated the estimation of immune cell composition, HLA expression levels, immune checkpoint expression, mutation burden, CSC index features, and the effectiveness of immunotherapy.ConclusionsThis study enhances our understanding of the role of PTFRs in CRC progression and introduces an innovative assessment framework for CRC immunotherapy. This framework improves the prediction of treatment outcomes and aids in the customization of therapeutic strategies.

The potential of a nomogram risk assessment model for the diagnosis of abdominal aortic aneurysm: a multicenter retrospective study

The incidence of abdominal aortic aneurysm (AAA) is very high, but there is no risk assessment model for early identification of AAA in clinic. The aim of this study was to develop a nomogram risk assessment model for predicting AAA. The data of 280 patients diagnosed as AAA and 385 controls in The Affiliated Suzhou Hospital of Nanjing Medical University were retrospectively reviewed. The LASSO regression method was applied to filter variables, and multivariate logistic regression was used to construct a nomogram. The discriminatory ability of the model was determined by calculating the area under the curve (AUC). The calibration capability of the model is evaluated by using bootstrap (resampling = 1000) internal validation and Hosmer–Lemeshow test. The clinical utility and clinical application value were evaluated by decision curve analysis (DCA) and clinical impact curve (CIC). In addition, a retrospective review of 133 AAA patients and 262 controls from The First Affiliated Hospital of Soochow University was performed as an external validation cohort. Eight variables are selected to construct the nomogram of AAA risk assessment model. The nomogram predicted AAA with AUC values of 0.928 (95%CI, 0.907–0.950) in the training cohort, and 0.902 (95%CI, 0.865–0.940) in the external validation cohort, the risk prediction model has excellent discriminative ability. The calibration curve and Hosmer–Lemeshow test proved that the nomogram predicted outcomes were close to the ideal curve, the predicted outcomes were consistent with the real outcomes, the DCA curve and CIC curve showed that patients could benefit. This finding was also confirmed in the external validation cohort. In this study, a nomogram was constructed that incorporated eight demographic and clinical characteristics of AAA patients, which can be used as a practical approach for the personalized early screening and auxiliary diagnosis of the potential risk factors.

Exploration of the ubiquitination-related molecular classification and signature to predict the survival and immune microenvironment in colon cancer.

Ubiquitination, a major post-translational modification, significantly impacts tumorigenesis, progression, and prognosis. This study aims to classify colon cancer at the molecular level and create a reliable signature using ubiquitination-related genes (URGs) to assess the immune microenvironment and prognosis. We employed non-negative matrix factorization to subtype colon cancer based on ubiquitination-related gene (URG) expression patterns. Quantitative scores for 28 immune cell infiltrates and the tumor microenvironment were computed using single-sample gene set enrichment analysis (ssGSEA) and the Estimate algorithm. Subtype feature genes were selected through Lasso logistic regression and SVM-RFE algorithm. The ubiquitination-related signature was constructed using univariate Cox, Lasso, and stepwise regression methods to categorize patients into high and low-risk groups. Validation included log-rank tests, receiver operating characteristic (ROC) analysis, decision curve analysis (DCA), and external dataset validation. Immune therapy response was compared using Tumor Immune Dysfunction and Exclusion (TIDE), Immunophenoscore (IPS), and submap analyses. Clinical variables and risk scores were integrated into an enhanced nomogram. The early diagnostic value of four URGs was confirmed via quantitative real-time polymerase chain reaction (qRT-PCR) and immunohistochemistry. The cell proliferation was assessed through colony formation, EdU staining, and xenograft tumorigenesis assays. Prognostic ubiquitination-related genes (URGs) stratified patients into subtypes, revealing differences in survival, immune cell infiltration, and pathological staging. A signature of 6 URGs (ARHGAP4, MID2, SIAH2, TRIM45, UBE2D2, WDR72) was identified from 57 subtype-related genes. The high-risk group exhibited characteristics indicative of enhanced epithelial-mesenchymal transition, immune escape, immunosuppressive myeloid-derived suppressor cells, regulatory T cell infiltration, and lower immunogenicity. In contrast, the low-risk group demonstrated the opposite trend but showed a better response to CTLA4 checkpoint inhibitors. The predictive performance of the nomogram significantly improved with the integration of risk score, stage, and age. ARHGAP4 and SIAH2 exhibit promising early diagnostic capabilities. Additionally, WDR72 knockdown significantly inhibited CRC cell proliferation both in vitro and in vivo. Our developed ubiquitination-related signature and genes serve as promising biomarkers for colon cancer prognosis, immune microenvironment, and diagnosis.

Predictive utility of emotional regulation abilities for assessing cognitive improvement in depression

ObjectiveTo construct a predictive model for the improvement of cognitive function in patients with depressive disorder treated with SNRIs, based on emotional regulation abilities, and to provide personalized treatment for depressed patients. MethodsClinical data from 170 patients with depressive disorder treated with SNRIs at Tongji Hospital, Shanghai, from December 2017 to May 2023 were collected. Based on whether the MoCA-B total score at 3–6 months post-treatment was at least 2 points higher than at baseline, patients were divided into the cognitive function improved group (n = 80) and the cognitive function not improved group (n = 90). Stepwise logistic regression and LASSO regression were used to select predictive factors, and logistic regression analysis was applied to construct predictive models solely based on emotional regulation abilities, combined with executive functions and HAMD scores. The models were further validated through Bootstrap internal validation, calibration curve plotting, and C-index calculation, and a comparison between the two models was performed. ResultsAn ER model with an area under the ROC curve of 0.817was established using four emotional regulation ability indicators: the valence of reappraised images, the arousal of negative images, the arousal of neutral images, and the success of reappraisal (arousal). Internal validation using Bootstrap showed a C index of 0.817, and clinical decision curves indicated that this model has a significant net benefit with a probability of improved cognitive function ranging from about 20 to 85%. Additionally, an EREH model including emotional regulation ability, executive function, and HAMD score as predictors was constructed using Lasso and logistic regression methods. This model reached an area under the ROC curve of 0.859and clinical decision curves showed high net benefits with probabilities of improved cognitive function ranging from 10 to 100%. The calibration curves of both models coincided well with the actual curves, with the latter having a higher AUC and significant statistical differences between the two models. ConclusionThis study suggests that emotional regulation ability may serve as a predictor for the improvement of cognitive functions in patients with depression depressive disorder treated with SNRIs. However, it is important to note that there may be other factors not covered or included in this study.The predictive model that includes executive functions and HAMD scores offers better differentiation and consistency and is more feasible in clinical practice.

Machine learning-based prediction of hazards fine PM2.5 concentrations: a case study of Delhi, India

The air quality of many geographical locations has been deteriorating in the last decades. This deterioration in air quality has affected a very large number of people, and they have been diagnosed with asthma and other respiratory problems. Among various pollutants, PM2.5 is the major cause of numerous health-related problems. Predicting PM2.5 concentration levels using ML models of these dissolved particles might help residents and the government prepare a better prevention and safety plan that can eventually lower the risk factor. The present study is based on predicting the PM2.5 concentration of Delhi by applying various ML models based on meteorological features like wind speed, temperature, humidity, visibility, etc. For prediction of PM2.5, linear regression, decision tree regression, RF regression, KNN regression, and Lasso regression methods were employed in the present study. The model performance was assessed using various parameters, including MAE, MSE, RMSE, and R2 Score. In a comparative study of all regression models, linear regression demonstrated the most favorable outcomes. The model exhibited a superior fit to the data, evidenced by its lowest RMSE value (52.19), outperforming the random forest regression (RMSE = 94.75), K Nearest Neighbor regression (RMSE = 83.93), each of which yielded higher RMSE scores compared to linear regression. Lasso regression (RMSE = 65.20) and decision tree (RMSE = 68.22) also exhibited improved performance following linear regression. The findings of this study advocate for implementing strategies to enforce stringent emission regulations for both industrial operations and vehicular activities. Such measures are imperative for mitigating air pollution levels and subsequently curtailing its adverse impacts on public health within the region. Additionally, this study underscores the necessity for further research endeavours to explore future avenues, with the aim of garnering global attention towards addressing this pressing issue.

Gene signatures of copper metabolism related genes may predict prognosis and immunity status in Ewing's sarcoma.

Cuproptosis is copper-induced cell death. Copper metabolism related genes (CMRGs) were demonstrated that used to assess the prognosis out of tumors. In the study, CMRGs were tested for their effect on TME cell infiltration in Ewing's sarcoma (ES). The GEO and ICGC databases provided the mRNA expression profiles and clinical features for downloading. In the GSE17674 dataset, 22prognostic-related copper metabolism related genes (PR-CMRGs) was identified by using univariate regression analysis. Subsequently, in order to compare the survival rates of groups with high and low expression of these PR-CMRGs,Kaplan-Meier analysis was implemented. Additionally, correlations among them were examined. The study employed functional enrichment analysis to investigate probable underlying pathways, while GSVA was applied to evaluate enriched pathways in the ES (Expression Set). Through an unsupervised clustering algorithm, samples were classified into two clusters, revealing significant differences in survival rates and levels of immune infiltration. Using Lasso and step regression methods, five genes (TFRC, SORD, SLC11A2, FKBP4, and AANAT) were selected as risk signatures. According to the Kaplan-Meier survival analysis, the high-risk group had considerably lower survival rates than the low-risk group(p=6.013e-09). The area under the curve (AUC) values for the receiver operating characteristic (ROC) curve were 0.876, 0.883, and 0.979 for 1, 3, and 5 years, respectively. The risk model was further validated in additional datasets, namely GSE63155, GSE63156, and the ICGC datasets. To aid in outcome prediction, a nomogram was developed that incorporated risk levels and clinical features. This nomogram's performance was effectively validated through calibration curves.Additionally, the study evaluated the variations in immune infiltration across different risk groups, as well as high-expression and low-expression groups. Importantly, several drugs were identified that displayed sensitivity, offering potential therapeutic options for ES. The findings above strongly indicate that CMRGs play crucial roles in predicting prognosis and immune status in ES.

Association between estimated plasma volume status and acute kidney injury in patients who underwent coronary revascularization: A retrospective cohort study from the MIMIC-IV database.

Acute kidney injury (AKI) remains a common complication of coronary revascularization and increases poor outcomes in critically ill surgical patients. Compared to the plasma volume status (PVS), estimated plasma volume status (ePVS) has the advantages of being noninvasive and simple and has been shown to be associated with worse prognosis in patients undergoing coronary revascularization. This study was to evaluate the association of ePVS with the risk of AKI in patients who underwent coronary revascularization. In this retrospective cohort study, data of patients who underwent coronary revascularization were extracted from the Medical Information Mart for Intensive Care (MIMIC)-IV database (2008-2019). The outcome was the occurrence of AKI after ICU admission. The covariates were screened via the LASSO regression method. Univariate and multivariate Logistic regression models were performed to assess the association of ePVS and PVS and the odds of AKI in patients who underwent coronary revascularization, with results shown as odds ratios (ORs) and 95% confidence intervals (CIs). Subgroup analyses of age, surgery, and anticoagulation agents and sequential organ failure assessment (SOFA) score were performed to further explore the association of ePVS with AKI. A total of 3,961 patients who underwent coronary revascularization were included in this study, of whom 2,863 (72.28%) had AKI. The high ePVS was associated with the higher odds of AKI in patients who received coronary revascularization (OR = 1.06, 95%CI: 1.02-1.10), after adjusting for the covariates such as age, race, SAPS-II score, SOFA score, CCI, weight, heart rate, WBC, RDW-CV, PT, BUN, glucose, calcium, PH, PaO2, mechanical ventilation, vasopressors, and diuretic. Similar results were found in patients who underwent the CABG (OR = 1.07, 95%CI: 1.02-1.11), without anticoagulation agents use (OR = 1.07, 95%CI: 1.03-1.12) and with high SOFA score (OR = 1.10, 95%CI: 1.04-1.17). No relationship was found between PVS and the odds of AKI in patients who underwent the coronary revascularization. The ePVS may be a promising parameter to evaluate the risk of AKI in patients undergoing coronary revascularization, which provides a certain reference for the risk stratification management of ICU patients who underwent coronary revascularization.

The impact of the combat method on radiomics feature compensation and analysis of scanners from different manufacturers

BackgroundThis study investigated whether the Combat compensation method can remove the variability of radiomic features extracted from different scanners, while also examining its impact on the subsequent predictive performance of machine learning models.Materials and methods135 CT images of Credence Cartridge Radiomic phantoms were collected and screened from three scanners manufactured by Siemens, Philips, and GE. 100 radiomic features were extracted and 20 radiomic features were screened according to the Lasso regression method. The radiomic features extracted from the rubber and resin-filled regions in the cartridges were labeled into different categories for evaluating the performance of the machine learning model. Radiomics features were divided into three groups based on the different scanner manufacturers. The radiomic features were randomly divided into training and test sets with a ratio of 8:2. Five machine learning models (lasso, logistic regression, random forest, support vector machine, neural network) were employed to evaluate the impact of Combat on radiomic features. The variability among radiomic features were assessed using analysis of variance (ANOVA) and principal component analysis (PCA). Accuracy, precision, recall, and area under the receiver curve (AUC) were used as evaluation metrics for model classification.ResultsThe principal component and ANOVA analysis results show that the variability of different scanner manufacturers in radiomic features was removed (P˃0.05). After harmonization with the Combat algorithm, the distributions of radiomic features were aligned in terms of location and scale. The performance of machine learning models for classification improved, with the Random Forest model showing the most significant enhancement. The AUC value increased from 0.88 to 0.92.ConclusionsThe Combat algorithm has reduced variability in radiomic features from different scanners. In the phantom CT dataset, it appears that the machine learning model’s classification performance may have improved after Combat harmonization. However, further investigation and validation are required to fully comprehend Combat’s impact on radiomic features in medical imaging.

Analysis of cell death-related genes to evaluate the prognostic and immunotherapeutic value in bladder cancer

To enhance therapeutic approaches, we created a distinctive pattern utilizing the cell demise indicator (CDI) to predict the effectiveness of immunotherapy in individuals with bladder carcinoma (BLCA). Hub prognostic CDIs were identified from the TCGA database using differential gene expression and survival analysis, encompassing 763 genes across 13 death modes. The subtype assessment was employed to evaluate the impact of these genes on the prognosis and immunotherapeutic outcomes in patients with BLCA. The LASSO regression method was used to identify significant CDIs, while Cox regression and nomogram analyses were conducted to explore the impact of CDIs on prognosis. CHMP4C and GSDMB were selected as the hub genes for the following research. Subsequently, These two central genes underwent further investigation to explore their association with immunotherapy, followed by an analysis of their potential regulatory network. Subtype analysis showed that these CDIs were significantly associated with the prognosis and immunotherapy of BLCA patients. The regulatory network in BLCA was evaluated through the establishment of the lncRNA XIST/NEAT1-CDIs-miR-146a-5p/miR-429 axis. Immunohistochemical analysis revealed a significant up-regulation of CHMP4C in bladder cancer tissues, which was strongly associated with an unfavorable prognosis for BLCA patients. Moreover, our findings provide compelling evidence that CHMP4C plays a pivotal role in promoting BLCA progression through the activation of the epithelial-mesenchymal transition (EMT) pathway. These findings highlight the negative impact of CHMP4C on BLCA patient prognosis, while also providing insights into the oncogenic mechanisms and immunotherapy in which CHMP4C may be involved.

Machine learning regression algorithms for predicting muscle, bone, carcass fat and commercial cuts in hairless lambs

The growth in demand and demand for quality in the sheep chain has generated the need for automation techniques in the meat industry and the need to obtain responses with greater speed and standardization. The research aimed to predict tissue characteristics of the carcass and commercial cuts based on measurements obtained by VIA – see oimage analysis, carried out on cold carcasses of hairless lambs, using machine learning employing regressive techniques for variable selection. Information from 72 carcasses of castrated male lambs, aged between 8 and 11 months, with an average cold carcass weight of 16.13 ± 3.98 kg, was used. Images of the right side of the carcasses were captured from the dorsal and lateral views using a digital camera. From the ImageJ2 software, VIA data, measurements and shape descriptors (areas, perimeters, widths, lengths, convexities, solidities) were obtained, combined with cold carcass weight and used to generate four sets of data, called descriptor sets (DSs). Obtaining DS1, DS1’, DS2, DS2’, DS3, DS3’, DS4 AND DS4’. To generate these sets, a database was formed and divided into a training bank (with 70% of the observations) and a test bank (30% of the observations). Multiple linear regression models were developed using Stepwise, LASSO, and Elastic Net regression methods, combined with k-fold cross-validation, to evaluate the performance of the models. The accuracy of the estimates was based on RMSE, R2, Pearson correlation and bias metrics. For the variables tested in this study, the proposed shape descriptors were mostly efficient in predicting tissue and weight variables. DS1' with the LASSO technique presented the best adjustments for variables total muscle and fat followed by shoulder, loin and rib cuts. The descriptors tested by this study were able to predict with quality the vast majority of the characteristics tested, the variable cold carcass weight (CCW), introduced as additional predictor, promoted a consistent improvement in the fits of all models. DS1 presented greater constancy for the twenty-three predicted characteristics and Stepwise presented the worst predictive performance, in relation to LASSO and Elastic Net. Despite close adjustments between the generated models, in general, Elastic Net presented lower performance than LASSO.

Risk prediction modelling in idiopathic inflammatory myositis-associated interstitial lung disease based on seven factors including serum KL-6 and lung ultrasound B-lines.

To develop a user-friendly nomogram-based predictive model for interstitial lung disease (ILD) in patients with idiopathic inflammatory myositis (IIM). A retrospective study was conducted at Shantou Central Hospital, encompassing 205 IIM patients diagnosed between January 2013 and December 2022. We used the LASSO regression method in the discovery set to select features for model construction, followed by efficacy verification through AUC of ROC. Afterwards, KL-6 values and LUS B-lines number were added into this model to evaluate whether these 2 factors added to the model efficiency. Finally, a web version was constructed to make it more available. Among the 205 IIM patients, 115 (56.1%) patients were diagnosed with ILD, and 90 (43.9%) did not. The predictive model, derived from the training set, comprised four independent risk factors, including age, presence of respiratory symptoms, anti-melanoma differentiation-associated gene 5 (MDA-5) antibody positivity, and anti-aminoacyl transfer RNA synthetase (anti-ARS) antibodies positivity. Notably, anti-TIF1-γ antibody positivity emerged as a protective factor. The AUC of the ROC based on these 5 factors was 0.876 in the training set and 0.861 in the validation set. The AUC of the ROC based on the 5 factors plus KL-6 was 0.922, 5 factors plus B-line number was 0.949 and 5 factors plus both KL-6 and B-line number was 0.951. Accordingly, a nomogram and a web version were developed. This predictive model demonstrates robust capability to assess ILD risk in IIM patients, particularly when augmented with serum KL-6 level or/and LUS B-line number.

Downscaling Future Precipitation over Mi Oya River Basin using Artificial Neural Networks

Studying future precipitation behaviour in river basins is essential for proper water resources and land-use planning within them, as this will help to reduce the risk and mitigate disasters that can occur in the future. General Circulation Models (GCMs) are used to study future precipitation fluctuations, which simulate large-scale climate variations under the effect of greenhouse gas changes. The GCM runs at a coarse spatial resolution which cannot be directly used for climate impact studies. Therefore, downscaling is required to extract the sub-grid and local scale information. This study examines the use of the Long Short-Term Memory (LSTM) neural network for climate downscaling to the Mi-Oya river basin in Sri Lanka using CNRM-CM5 and HadCM3 GCMs and observed annual data for 35 years. The precipitation data were extracted to cover Sri Lanka. Current downscaling models mostly use Convolutional Neural Networks (CNNs) to downscale GCMs. Out of 42 GCMs, two appropriate GCMs were chosen using the data analysis tool Data Integration and Analysis System (DIAS). The best predictor variables were chosen using the LASSO regression method. In this research, Machine Learning models were implemented using the Google TensorFlow platform. The Nash–Sutcliffe coefficient, Pearson correlation coefficient, and root-mean-square error performance indices were used to evaluate the performances of different downscaling models. Statistical downscaling was performed on the data at RCP 2.6, 4.5, and 8.5 using a LSTM. Subsequently, the changes that would take place by the year 2100 were analysed. The results show that precipitation will be reduced in the 2nd and 3rd decades of the 21st century, and precipitation will increase toward the 22nd century.

Optimal Decision on Pricing and Replenishment of Vegetable Products Based on Goal Planning

Developing replenishment and pricing plans has always been the key for fresh food supermarkets to gain more profits. This paper analyzes various situations and provides the optimal solution through Spearman correlation analysis, Lasso regression, goal programming and other methods and models. We use normality testing and box plots to present the distribution patterns of major categories and individual products from both vertical and horizontal perspectives. Establish a single objective programming model and solve it with Lingo to obtain a maximum profit of 98513 yuan for 7 days. The largest total replenishment amount is for the flower and leaf category, the smallest is for edible mushrooms, the highest average cost markup rate is for aquatic roots and stems, and the lowest is for cauliflower. This has certain reference value and theoretical significance for fresh food supermarkets to formulate replenishment and pricing plans.

An evaluation of variable selection methods using Southern Africa solar irradiation data

Dimensionality poses a challenge in developing quality predictive models. Often when modelling solar irradiance (SI), many covariates are considered. Training such data has several disadvantages. This study sought to identify the best variable embedded selection method for different location and time horizon combinations from Southern Africa solar irradiance data. It introduced new variable selection methods into solar irradiation studies, namely penalised quantile regression (PQR), regularised random forests (RRF), and quantile regression forest (QRF). Stability analysis, performance and accuracy metric evaluations were used to compare them with the common lasso, elastic and ridge regression methods. The QRF model performed best in all locations followed by the shrinkage methods on hourly data. However, it was found that QRF is not sensitive to associations through correlations, thereby ignoring the relevance of variables while focusing on importance. Among the shrinkage methods, the lasso performed best in only one location. On the 24-hour horizon, elastic net dominated the performances among the shrinkage methods, but QRF was best in three locations of the six considered. Results confirmed that variable selection methods performed differently on different situational data sets. Depending on the strengths of the methods, results were combined to identify the most paramount variables. Day, total rainfall, and wind direction were superfluous features in all situations. The study concluded that shrinkage methods are best in cases of extreme multicollinearity, while QRF is best on data sets with outliers or/and heavy tails.