Intelligence analysis of membrane distillation via machine learning models for pharmaceutical separation.

This study investigates simulation of pharmaceutical separation via membrane distillation process by computational simulation and machine learning modeling strategy. The efficacy of three regression models, i.e., Multi-layer Perceptron (MLP), Gamma Regression, and Support Vector Regression (SVR) in predicting the solute concentration, C(mol/m³), was evaluated. The hyper-parameters were optimized by fine-tuning the models using the Red Deer Algorithm (RDA). Computational analyses were carried out for removal of pharmaceuticals from solution by membrane distillation in continuous mode. Mass transfer and machine learning models were implemented focusing on concentration of solute in the feed section of membrane. Results indicate that the Multi-layer Perceptron model achieved great accuracy with an R2 of 0.9955, an MAE of 0.0084, and an RMSE of 0.0148, effectively capturing complex nonlinear relationships in the data. Gamma Regression also performed acceptably, with fitting R2 of 0.9214, showing its suitability for positively skewed data. The Support Vector Regression model, while capturing the general trend, showed the lowest performance with an R2 of 0.8710. These findings suggest that the Multi-layer Perceptron is the most accurate model for this dataset, followed by Gamma Regression and Support Vector Regression. This underscores the importance of careful model selection and optimization in regression analysis in combination with computational simulation of membrane processes.

A novel Skin lesion prediction and classification technique: ViT-GradCAM.

BACKGROUND: Skin cancer is one of the highly occurring diseases in human life. Early detection and treatment are the prime and necessary points to reduce the malignancy of infections. Deep learning techniques are supplementary tools to assist clinical experts in detecting and localizing skin lesions. Vision transformers (ViT) based on image segmentation classification using multiple classes provide fairly accurate detection and are gaining more popularity due to legitimate multiclass prediction capabilities. MATERIALS AND METHODS: In this research, we propose a new ViT Gradient-Weighted Class Activation Mapping (GradCAM) based architecture named ViT-GradCAM for detecting and classifying skin lesions by spreading ratio on the lesion's surface area. The proposed system is trained and validated using a HAM 10000 dataset by studying seven skin lesions. The database comprises 10 015 dermatoscopic images of varied sizes. The data preprocessing and data augmentation techniques are applied to overcome the class imbalance issues and improve the model's performance. RESULT: The proposed algorithm is based on ViT models that classify the dermatoscopic images into seven classes with an accuracy of 97.28%, precision of 98.51, recall of 95.2%, and an F1 score of 94.6, respectively. The proposed ViT-GradCAM obtains better and more accurate detection and classification than other state-of-the-art deep learning-based skin lesion detection models. The architecture of ViT-GradCAM is extensively visualized to highlight the actual pixels in essential regions associated with skin-specific pathologies. CONCLUSION: This research proposes an alternate solution to overcome the challenges of detecting and classifying skin lesions using ViTs and GradCAM, which play a significant role in detecting and classifying skin lesions accurately rather than relying solely on deep learning models.

Artificial neural network based forecasting of diesel engine performance and emissions utilizing waste cooking biodiesel.

Ecological and environmental problems resulting from fossil fuels are due to the harmful emissions released into the atmosphere. The rising interest in searching for alternative fuels like biodiesel is growing to solve these problems. Waste cooking oil (WCO) is transformed into methyl ester and combined with biodiesel in percentages of 25, 50, 75, and 100%. Research is done on the impacts of methyl ester blends on engine performance and emissions. Compared to diesel, the methyl ester combination showed 25% lower brake power and 24% loss in thermal efficiency at maximum load and 1500 rpm. However, diesel fuel showed 23% lower specific fuel consumption increase than biodiesel. Compared to diesel, methyl ester exhibits 15% lower air-fuel ratio and 4% volumetric efficiency. Biodiesel lowers CO, HC, and smoke concentrations by 12, 44, and 48%, respectively, compared to diesel. Biodiesel emits 23% higher NOx at 1500 rpm and 100% engine load. To predict the emissions and performance of different percentages of biodiesel at engine speed variation, an artificial neural network (ANN) model is presented. ANN modeling minimizes labor, time, and finances and uses nonlinear data. Predictions were produced about the brake output power, specific fuel consumption, thermal efficiency, air-fuel ratio, volumetric efficiency, and emissions of smoke, CO, HC, and NOx as a function of engine speed and blend ratio. All correlation coefficients (r) over 0.99 and R 2 values were beyond 0.98 for all variables. There were low values of MSE, MAPE, and MSLE with significant predictive ability. WCO's biodiesel is a viable diesel engine replacement fuel.

Application of a Multi-Layer Perceptron and Markov Chain Analysis-Based Hybrid Approach for Predicting and Monitoring LULCC Patterns Using Random Forest Classification in Jhelum District, Punjab, Pakistan.

Land-use and land-cover change (LULCC) is a critical environmental issue that has significant effects on biodiversity, ecosystem services, and climate change. This study examines the land-use and land-cover (LULC) spatiotemporal dynamics across a three-decade period (1998-2023) in a district area. In order to forecast the LULCC patterns, this study suggests a hybrid strategy that combines the random forest method with multi-layer perceptron (MLP) and Markov chain analysis. To predict the dynamics of LULC changes for the year 2035, a hybrid technique based on multi-layer perceptron and Markov chain model analysis (MLP-MCA) was employed. The area of developed land has increased significantly, while the amount of bare land, vegetation, and forest cover have all decreased. This is because the principal land types have changed due to population growth and economic expansion. This study also discovered that between 1998 and 2023, the built-up area increased by 468 km2 as a result of the replacement of natural resources. It is estimated that 25.04% of the study area's urbanization will increase by 2035. The performance of the model was confirmed with an overall accuracy of 90% and a kappa coefficient of around 0.89. It is important to use advanced predictive models to guide sustainable urban development strategies. The model provides valuable insights for policymakers, land managers, and researchers to support sustainable land-use planning, conservation efforts, and climate change mitigation strategies.

Human lung cancer classification and comprehensive analysis using different machine learning techniques.

Lung cancer is the most common causes of death among all cancer-related diseases. A lung scan examination of the patient is the primary diagnostic technique. This scan analysis pertains to an MRI, CT, or X-ray. The automated classification of lung cancer is difficult due to the involvement of multiple steps in imaging patients' lungs. In this manuscript, human lung cancer classification and comprehensive analysis using different machine learning techniques is proposed. Initially, the input images are gathered using lung cancer dataset. The proposed method processes these images using image-processing techniques, and further machine learning techniques are utilized for categorization. Seven different classifiers including the k-nearest neighbors (KNN), support vector machine (SVM), decision tree (DT), multinomial naive Bayes (MNB), stochastic gradient descent (SGD), random forest (RF), and multi-layer perceptron (MLP) classifier are used, which classifies the lung cancer as malignant and benign. The performance of the proposed approach is examined using performances metrics, like positive predictive value, accuracy, sensitivity, and f-score are evaluated. Among them, the performance of the MLP classifier provides 25.34%, 45.39%, 15.39%, 41.28%, 22.17%, and 12.12% higher accuracy than other KNN, SVM, DT, MNB, SGD, and RF respectively. RESEARCH HIGHLIGHTS: Lung cancer is a leading cause of cancer-related death. Imaging (MRI, CT, and X-ray) aids diagnosis. Automated classification of lung cancer faces challenges due to complex imaging steps. This study proposes human lung cancer classification using diverse machine learning techniques. Input images from lung cancer dataset undergo image processing and machine learning. Classifiers like k-nearest neighbors, support vector machine, decision tree, multinomial naive Bayes, stochastic gradient descent, random forest, and multi-layer perceptron (MLP) classify cancer types; MLP excels in accuracy.

Geographic inequities in neonatal survival in Nigeria: a cross-sectional evidence from spatial and artificial neural network analyses.

This study was conducted to provide empirical evidence of geographical variations of neonatal mortality and its associated social determinants with a view to improving neonatal survival at the subnational level in Nigeria. With a combination of spatial analysis and artificial intelligence techniques, this study analysed data from the 2016/2017 Nigeria Multiple Indicator Cluster Survey. The analysis focused on the neonatal period of a weighted national representative population of 30,924 live births delivered five years before the survey commencement. Global Moran's I index and local indicator of spatial autocorrelation cluster maps were used to determine hot and cold spots. A multilayer perceptron neural network was used to identify the key determinants of neonatal mortality across the states and geopolitical zones in Nigeria. The overall neonatal mortality rate was 38 deaths per 1000 live births. There is evidence of geographic clustering of neonatal mortality across Nigeria (worse in the North-Central and North-West zones), majorly driven by poor maternal access to mass media (which plays a critical role in promoting positive health behaviours), short birth interval, a higher position in a family birth order, and young maternal age at child's birth. This study highlights the need for a policy shift towards implementing state and region-specific strategies in Nigeria. Gender-responsive, culturally, and regionally appropriate reproductive, maternal, and child health-targeted interventions may address geographical inequity in neonatal survival.

Predicting the Association of Metabolites with Both Pathway Categories and Individual Pathways.

Metabolism is a network of chemical reactions that sustain cellular life. Parts of this metabolic network are defined as metabolic pathways containing specific biochemical reactions. Products and reactants of these reactions are called metabolites, which are associated with certain human-defined metabolic pathways. Metabolic knowledgebases, such as the Kyoto Encyclopedia of Gene and Genomes (KEGG) contain metabolites, reactions, and pathway annotations; however, such resources are incomplete due to current limits of metabolic knowledge. To fill in missing metabolite pathway annotations, past machine learning models showed some success at predicting the KEGG Level 2 pathway category involvement of metabolites based on their chemical structure. Here, we present the first machine learning model to predict metabolite association to more granular KEGG Level 3 metabolic pathways. We used a feature and dataset engineering approach to generate over one million metabolite-pathway entries in the dataset used to train a single binary classifier. This approach produced a mean Matthews correlation coefficient (MCC) of 0.806 ± 0.017 SD across 100 cross-validation iterations. The 172 Level 3 pathways were predicted with an overall MCC of 0.726. Moreover, metabolite association with the 12 Level 2 pathway categories was predicted with an overall MCC of 0.891, representing significant transfer learning from the Level 3 pathway entries. These are the best metabolite pathway prediction results published so far in the field.

Reliability-Based Design Optimization for Polymer Electrolyte Membrane Fuel Cells: Tackling Dimensional Uncertainties in Manufacturing and Their Effects on Costs of Cathode Gas Diffusion Layer and Bipolar Plates.

Polymer Electrolyte Membrane Fuel Cells (PEMFCs) have emerged as a pivotal technology in the automotive industry, significantly contributing to the reduction of greenhouse gas emissions. However, the high material costs of the gas diffusion layer (GDL) and bipolar plate (BP) create a barrier for large scale commercial application. This study aims to address this challenge by optimizing the material and design of the cathode, GDL and BP. While deterministic design optimization (DDO) methods have been extensively studied, they often fall short when manufacturing uncertainties are introduced. This issue is addressed by introducing reliability-based design optimization (RBDO) to optimize four key PEMFC design variables, i.e., gas diffusion layer thickness, channel depth, channel width and land width. The objective is to maximize cell voltage considering the material cost of the cathode gas diffusion layer and cathode bipolar plate as reliability constraints. The results of the DDO show an increment in cell voltage of 31 mV, with a reliability of around 50% in material cost for both the cathode GDL and cathode BP. In contrast, the RBDO method provides a reliability of 95% for both components. Additionally, under a high level of uncertainty, the RBDO approach reduces the material cost of the cathode GDL by up to 12.25 $/stack, while the material cost for the cathode BP increases by up to 11.18 $/stack Under lower levels of manufacturing uncertainties, the RBDO method predicts a reduction in the material cost of the cathode GDL by up to 4.09 $/stack, with an increase in the material cost for the cathode BP by up to 6.71 $/stack, while maintaining a reliability of 95% for both components. These results demonstrate the effectiveness of the RBDO approach in achieving a reliable design under varying levels of manufacturing uncertainties.

Comprehensive input models and machine learning methods to improve permeability prediction.

This study investigates the use of machine learning techniques and the proper selection of input data to estimate permeability in geosciences, using six types of input logs: gamma ray (GR), resistivity (RT), effective porosity (PHIE), density (RHO), sonic (DT), and compensated neutron porosity (NPHI). A total of 57 models were constructed using combinations of these logs and tested using five machine learning methods: Extreme Learning Machine (ELM), Random Forest (RF), Gradient Boosting (GB), K-Nearest Neighbor (KNN), and Multilayer Perceptron (MLP). This approach produced 285 unique permeability predictions. RF had the highest correlation coefficient (0.925) and average error (0.196), indicating a precision-correlation trade-off. The ELM approach had the lowest average error, 0.083, and a correlation value of 0.871. Testing on a blind well revealed that the GB and RF approaches were highly effective in predicting permeability, with R² values of 0.92 and 0.90, respectively, even in untested settings. The findings emphasize the need of using appropriate machine learning algorithms and input data to improve model accuracy and reliability.

Choosing an appropriate somatic embryogenesis medium of carrot (Daucus carota L.) by data mining technology.

INTRODUCTION: Developing somatic embryogenesis is one of the main steps in successful in vitro propagation and gene transformation in the carrot. However, somatic embryogenesis is influenced by different intrinsic (genetics, genotype, and explant) and extrinsic (e.g., plant growth regulators (PGRs), medium composition, and gelling agent) factors which cause challenges in developing the somatic embryogenesis protocol. Therefore, optimizing somatic embryogenesis is a tedious, time-consuming, and costly process. Novel data mining approaches through a hybrid of artificial neural networks (ANNs) and optimization algorithms can facilitate modeling and optimizing in vitro culture processes and thereby reduce large experimental treatments and combinations. Carrot is a model plant in genetic engineering works and recombinant drugs, and therefore it is an important plant in research works. Also, in this research, for the first time, embryogenesis in carrot (Daucus carota L.) using Genetic algorithm (GA) and data mining technology has been reviewed and analyzed. MATERIALS AND METHODS: In the current study, data mining approach through multilayer perceptron (MLP) and radial basis function (RBF) as two well-known ANNs were employed to model and predict embryogenic callus production in carrot based on eight input variables including carrot cultivars, agar, magnesium sulfate (MgSO4), calcium dichloride (CaCl2), manganese (II) sulfate (MnSO4), 2,4-dichlorophenoxyacetic acid (2,4-D), 6-benzylaminopurine (BAP), and kinetin (KIN). To confirm the reliability and accuracy of the developed model, the result obtained from RBF-GA model were tested in the laboratory. RESULTS: The results showed that RBF had better prediction efficiency than MLP. Then, the developed model was linked to a genetic algorithm (GA) to optimize the system. To confirm the reliability and accuracy of the developed model, the result of RBF-GA was experimentally tested in the lab as a validation experiment. The result showed that there was no significant difference between the predicted optimized result and the experimental result. CONCLUTIONS: Generally, the results of this study suggest that data mining through RBF-GA can be considered as a robust approach, besides experimental methods, to model and optimize in vitro culture systems. According to the RBF-GA result, the highest somatic embryogenesis rate (62.5%) can be obtained from Nantes improved cultivar cultured on medium containing 195.23 mg/l MgSO4, 330.07 mg/l CaCl2, 18.3 mg/l MnSO4, 0.46 mg/l 2,4- D, 0.03 mg/l BAP, and 0.88 mg/l KIN. These results were also confirmed in the laboratory.

Coupling artificial neural network and sperm swarm optimization for soil temperature prediction at multiple depths.

Soil temperature (ST) stands as a pivotal parameter in the realm of water resources and irrigation. It serves as a guide for farmers, enabling them to determine optimal planting and fertilization timings. In the backdrop of regions like Iran, where water resources are scarce, a proficient and economical prediction model for ST, particularly at lower depths, becomes imperative. While recent models have demonstrated adeptness in predicting ST, in general, their error decreases with increasing depth, so that they had the lowest error at a depth of 100 cm. Addressing this gap, our study pioneers a novel hybrid model that excels in accurate daily ST prediction as it delves deeper. The models deployed encompass the multilayer perceptron (MLP) and an enhanced version, MLP coupled with the Sperm Swarm Optimization Algorithm (MLP-SSO). These models prognosticate daily ST across varying depths (5-100 cm), leveraging meteorological parameters such as air temperature, relative humidity, wind speed, sunshine hours, and precipitation. These parameters are anchored to the Ahvaz and Sabzevar synoptic stations in Iran, spanned over the period from 1997 to 2022. Evaluation of our research outcomes unveils that the root mean square error (RMSE) witnesses its most substantial reduction at a depth of 100 cm. For instance, at the Ahvaz station, the MLP-SSO model diminishes the RMSE value from 1.25 to 1.12 °C, in contrast to the MLP model. Similarly, at the Sabzevar station, the RMSE value drops from 1.78 to 1.49 °C using the coupled MLP-SSO model. These results robustly highlight the considerable enhancement brought about by the utilization of the MLP-SSO model, clearly surpassing the performance of the standalone MLP model. This emphasizes the potential and promise of the MLP-SSO model for future investigations, offering insights that can significantly advance the domain of soil temperature prediction and its implications for agricultural decision-making.

Genetic algorithm multiple linear regression and machine learning-driven QSTR modeling for the acute toxicity of sterol biosynthesis inhibitor fungicides.

Sterol Biosynthesis Inhibitors (SBIs) are a major class of fungicides used globally. Their widespread application in agriculture raises concerns about potential harm and toxicity to non-target organisms, including humans. To address these concerns, a quantitative structure-toxicity relationship (QSTR) modeling approach has been developed to assess the acute toxicity of 45 different SBIs. The genetic algorithm (GA) was used to identify key molecular descriptors influencing toxicity. These descriptors were then used to build robust QSTR models using multiple linear regression (MLR), support vector regression (SVR), and artificial neural network (ANN) algorithms. The Cross-validation, Y-randomization test, applicability domain methods, and external validation were carried out to evaluate the accuracy and validity of the generated models. The MLR model exhibited satisfactory predictive performance, with an R2 of 0.72. The SVR and ANN models obtained R2 values of 0.7 and 0.8, respectively. ANN model demonstrated superior performance compared to other models, achieving R2 cv and R2 test values of 0.74 and 0.7, respectively. The models passed both internal and external validation, indicating their robustness. These models offer a valuable tool for risk assessment, enabling the evaluation of potential hazards associated with future applications of SBIs.

Sex-specific Prediction of Cardiogenic Shock After Acute Coronary Syndromes: The SEX-SHOCK Score.

BACKGROUND: and aims: Cardiogenic shock (CS) remains the primary cause of in-hospital death after acute coronary syndromes (ACS), with its plateauing mortality rates approaching 50%. To test novel interventions, personalized risk prediction is essential. The ORBI (Observatoire Régional Breton sur l'Infarctus) score represents the first-of-its-kind risk score to predict in-hospital CS in ACS patients undergoing percutaneous coronary intervention (PCI). However, its sex-specific performance remains unknown, and refined risk prediction strategies are warranted. METHODS: This multinational study included a total of 53 537 ACS patients without CS on admission undergoing PCI. Following sex-specific evaluation of ORBI, regression and machine-learning models were used for variable selection and risk prediction. By combining best-performing models with highest-ranked predictors, SEX-SHOCK was developed, and internally and externally validated. RESULTS: The ORBI score showed lower discriminative performance for the prediction of CS in females than males in Swiss (AUC [95% CI]: 0.78 [0.76-0.81] vs. 0.81 [0.79-0.83]; p=0.048) and French ACS patients (0.77 [0.74-0.81] vs. 0.84 [0.81-0.86]; p=0.002). The newly developed SEX-SHOCK score, now incorporating ST-segment elevation, creatinine, C-reactive protein, and left ventricular ejection fraction, outperformed ORBI in both sexes (females: 0.81 [0.78-0.83]; males: 0.83 [0.82-0.85]; p<0.001), which prevailed following internal and external validation in RICO (females: 0.82 [0.79-0.85]; males: 0.88 [0.86-0.89]; p<0.001) and SPUM-ACS (females: 0.83 [0.77-0.90], p=0.004; males: 0.83 [0.80-0.87], p=0.001). CONCLUSIONS: The ORBI score showed modest sex-specific performance. The novel SEX-SHOCK score provides superior performance in females and males across the entire spectrum of ACS, thus providing a basis for future interventional trials and contemporary ACS management.

Microcystis abundance is predictable through ambient bacterial communities: A data-oriented approach.

The number of cyanobacterial harmful algal blooms (cyanoHABs) has increased, leading to the widespread development of prediction models for cyanoHABs. Although bacteria interact closely with cyanobacteria and directly affect cyanoHABs occurrence, related modeling studies have rarely utilized microbial community data compared to environmental data such as water quality. In this study, we built a machine learning model, the multilayer perceptron (MLP), for the prediction of Microcystis dynamics using both bacterial community and weekly water quality data from the Daechung Reservoir and Nakdong River, South Korea. The modeling performance, indicated by the R2 value, improved to 0.97 in the model combining bacterial community data with environmental factors, compared to 0.78 in the model using only environmental factors. This underscores the importance of microbial communities in cyanoHABs prediction. Through the post-hoc analysis of the MLP models, we revealed that nitrogen sources played a more critical role than phosphorus sources in Microcystis blooms, whereas the bacterial amplicon sequence variants did not have significant differences in their contribution to each other. Similar to the MLP model results, bacterial data also had higher predictability in multiple linear regression (MLR) than environmental data. In both the MLP and MLR models, Microscillaceae showed the strongest association with Microcystis. This modeling approach provides a better understanding of the interactions between bacteria and cyanoHABs, facilitating the development of more accurate and reliable models for cyanoHABs prediction using ambient bacterial data.

XGBoost algorithm assisted multi-component quantitative analysis with Raman spectroscopy.

To improve prediction performance and reduce artifacts in Raman spectra, we developed an eXtreme Gradient Boosting (XGBoost) preprocessing method to preprocess the Raman spectra of glucose, glycerol and ethanol mixtures. To ensure the robustness and reliability of the XGBoost preprocessing method, an X-LR model (which combined XGBoost preprocessing and a linear regression (LR) model) and a X-MLP model (which combined XGBoost preprocessing and a multilayer perceptron (MLP) model) were developed. These two models were used to quantitatively analyze the concentrations of glucose, glycerol and ethanol in the Raman spectra of mixed solutions. The proportion map of hyperparameters was firstly used to narrow down the search range of hyperparameters in the X-LR and the X-MLP models. Then the correlation coefficients (R2), root mean square of calibration (RMSEC), and root mean square error of prediction (RMSEP) were used to evaluate the models' performance. Experimental results indicated that the XGBoost preprocessing method achieved higher accuracy and generalization capability, with better performance than those of other preprocessing methods for both LR and MLP models.

Resolving autism spectrum disorder (ASD) through brain topologies using fMRI dataset with multi-layer perceptron (MLP).

Autism is a neurodevelopmental disorder that manifests in individuals during childhood and has enduring consequences for their social interactions and communication. The prediction of Autism Spectrum Disorder (ASD) in individuals based on the differences in brain networks and activities have been studied extensively in the recent past, however, with lower accuracies. Therefore in this research, identification at the early stage through computer-aided algorithms to differentiate between ASD and TD patients is proposed. In order to identify features, a Multi-Layer Perceptron (MLP) model is developed which utilizes logistic regression on characteristics extracted from connectivity matrices of subjects derived from fMRI images. The features that significantly contribute to the classification of individuals as having Autism Spectrum Disorder (ASD) or typically developing (TD) are identified by the logistic regression model. To enhance emphasis on essential attributes, an AND operation is integrated. This involves selecting features demonstrating statistical significance across diverse logistic regression analyses conducted on various random distributions. The iterative approach contributes to a comprehensive understanding of relevant features for accurate classification. By implementing this methodology, the estimation of feature importance became more dependable, and the potential for overfitting is moderated through the evaluation of model performance on various subsets of data. It is observed from the experimentation that the highly correlated Left Lateral Occipital Cortex and Right Lateral Occipital Cortex ROIs are only found in ASD. Also, it is noticed that the highly correlated Left Cerebellum Tonsil and Right Cerebellum Tonsil are only found in TD participants. Among the MLP classifier, a recall of 82.61 % is achieved followed by Logistic Regression with an accuracy of 72.46 %. MLP also stands out with a commendable accuracy of 83.57 % and AUC of 0.978.

Predicting the Pathway Involvement of Metabolites in Both Pathway Categories and Individual Pathways.

Metabolism is the network of chemical reactions that sustain cellular life. Parts of this metabolic network are defined as metabolic pathways containing specific biochemical reactions. Products and reactants of these reactions are called metabolites, which are associated with certain human-defined metabolic pathways. Metabolic knowledgebases, such as the Kyoto Encyclopedia of Gene and Genomes (KEGG) contain metabolites, reactions, and pathway annotations; however, such resources are incomplete due to current limits of metabolic knowledge. To fill in missing metabolite pathway annotations, past machine learning models showed some success at predicting KEGG Level 2 pathway category involvement of metabolites based on their chemical structure. Here, we present the first machine learning model to predict metabolite association to more granular KEGG Level 3 metabolic pathways. We used a feature and dataset engineering approach to generate over one million metabolite-pathway entries in the dataset used to train a single binary classifier. This approach produced a mean Matthews correlation coefficient (MCC) of 0.806 ± 0.017 SD across 100 cross-validations iterations. The 172 Level 3 pathways were predicted with an overall MCC of 0.726. Moreover, metabolite association with the 12 Level 2 pathway categories were predicted with an overall MCC of 0.891, representing significant transfer learning from the Level 3 pathway entries. These are the best metabolite-pathway prediction results published so far in the field.

Predicting hospitalization costs for pulmonary tuberculosis patients based on machine learning.

BACKGROUND: Pulmonary tuberculosis (PTB) is a prevalent chronic disease associated with a significant economic burden on patients. Using machine learning to predict hospitalization costs can allocate medical resources effectively and optimize the cost structure rationally, so as to control the hospitalization costs of patients better. METHODS: This research analyzed data (2020-2022) from a Kashgar pulmonary hospital's information system, involving 9570 eligible PTB patients. SPSS 26.0 was used for multiple regression analysis, while Python 3.7 was used for random forest regression (RFR) and MLP. The training set included data from 2020 and 2021, while the test set included data from 2022. The models predicted seven various costs related to PTB patients, including diagnostic cost, medical service cost, material cost, treatment cost, drug cost, other cost, and total hospitalization cost. The model's predictive performance was evaluated using R-square (R2), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE) metrics. RESULTS: Among the 9570 PTB patients included in the study, the median and quartile of total hospitalization cost were 13,150.45 (9891.34, 19,648.48) yuan. Nine factors, including age, marital status, admission condition, length of hospital stay, initial treatment, presence of other diseases, transfer, drug resistance, and admission department, significantly influenced hospitalization costs for PTB patients. Overall, MLP demonstrated superior performance in most cost predictions, outperforming RFR and multiple regression; The performance of RFR is between MLP and multiple regression; The predictive performance of multiple regression is the lowest, but it shows the best results for Other costs. CONCLUSION: The MLP can effectively leverage patient information and accurately predict various hospitalization costs, achieving a rationalized structure of hospitalization costs by adjusting higher-cost inpatient items and balancing different cost categories. The insights of this predictive model also hold relevance for research in other medical conditions.

WI-TMLEGA: Weight Initialization and Training Method Based on Entropy Gain and Learning Rate Adjustment.

Addressing the issues of prolonged training times and low recognition rates in large model applications, this paper proposes a weight training method based on entropy gain for weight initialization and dynamic adjustment of the learning rate using the multilayer perceptron (MLP) model as an example. Initially, entropy gain was used to replace random initial values for weight initialization. Subsequently, an incremental learning rate strategy was employed for weight updates. The model was trained and validated using the MNIST handwritten digit dataset. The experimental results showed that, compared to random initialization, the proposed initialization method improves training effectiveness by 39.8% and increases the maximum recognition accuracy by 8.9%, demonstrating the feasibility of this method in large model applications.

Machine Learning Techniques for Blind Beam Alignment in mmWave Massive MIMO.

This paper proposes methods for Machine Learning (ML)-based Beam Alignment (BA), using low-complexity ML models, and achieves a small pilot overhead. We assume a single-user massive mmWave MIMO, Uplink, using a fully analog architecture. Assuming large-dimension codebooks of possible beam patterns at UE and BS, this data-driven and model-based approach aims to partially and blindly sound a small subset of beams from these codebooks. The proposed BA is blind (no CSI), based on Received Signal Energies (RSEs), and circumvents the need for exhaustively sounding all possible beams. A sub-sampled subset of beams is then used to train several ML models such as low-rank Matrix Factorization (MF), non-negative MF (NMF), and shallow Multi-Layer Perceptron (MLP). We provide an extensive mathematical description of these models and the algorithms for each of them. Our extensive numerical results show that, by sounding only 10% of the beams from the UE and BS codebooks, the proposed ML tools are able to accurately predict the non-sounded beams through multiple transmitted power regimes. This observation holds as the codebook sizes at UE and BS vary from 128×128 to 1024×1024.