LightGBM Model Research Articles

Prediction of risk of in-hospital death in patients with chronic heart failure complicated by lung infections using interpretable machine learning

To predict the risk of in-hospital death in patients with chronic heart failure (CHF) complicated by lung infections using interpretable machine learning. The clinical data of 1415 patients diagnosed with CHF complicated by lung infections were obtained from the MIMIC-IV database. According to the pathogen type, the patients were categorized into bacterial pneumonia and non-bacterial pneumonia groups, and their risks of in-hospital death were compared using Kaplan-Meier survival curves. Univariate analysis and LASSO regression were used to select the features for constructing LR, AdaBoost, XGBoost, and LightGBM models, and their performance was compared in terms of accuracy, precision, F1 value, and AUC. External validation of the models was performed using the data from eICU-CRD database. SHAP algorithm was applied for interpretive analysis of XGBoost model. Among the 4 constructed models, the XGBoost model showed the highest accuracy and F1 value for predicting the risk of in-hospital death in CHF patients with lung infections in the training set. In the external test set, the XGBoost model had an AUC of 0.691 (95% CI: 0.654-0.720) in bacterial pneumonia group and an AUC of 0.725 (95% CI: 0.577-0.782) in non-bacterial pneumonia group, and showed better predictive ability and stability than the other models. The overall performance of the XGBoost model is superior to the other 3 models for predicting the risk of in-hospital death in CHF patients with lung infections. The SHAP algorithm provides a clear interpretation of the model to facilitate decision-making in clinical settings.

A light gradient boosting machine-based method for predicting the dynamic response of functionally graded plates

The primary objective of this paper is to efficiently predict the dynamic response of functionally graded plates using LightGBM – a light gradient boosting machine, without reliance on supplementary analysis tools. To obtain the optimal LightGBM model, a dataset comprising 1,000 pairs of input and output is generated through iterations using a combination of isogeometric analysis (IGA) and third-order shear deformation plate theory (TSDT). In this model, the input is represented by a power index which governs the material distribution of the plate, and the output comprises 200 values illustrating deflection over time. To demonstrate the effectiveness of LightGBM in terms of accuracy and computational time, the results obtained by the proposed model are compared to those achieved with the optimal ANN, XGBoost models, and IGA. ® 2024 Journal of Science and Technology - NTTU

A stacking ensemble model for predicting soil organic carbon content based on visible and near-infrared spectroscopy

The content of soil organic carbon (SOC) plays an important role in maintaining ecosystem functions, protecting soil biodiversity, and understanding carbon cycling processes. The combination of visible and near-infrared spectroscopy (VIS–NIRS) and machine learning can achieve rapid prediction of SOC content. However, it is still relatively unknown how to integrate the characteristics of various machine learning models to improve the performance of SOC prediction models. In this study, a new model for predicting SOC content based on stacking ensemble learning was proposed by using VIS–NIRS. The prediction performances of six different models including Support Vector Regression (SVR), Extreme Gradient Boosting (XGBoost), Random Forest (RF), Light Gradient Boosting Machine (LightGBM), PartialLeast-square (PLS) and Extreme Learning Machine (ELM) on SOC content under different spectral preprocessing methods were compared. The results indicated that SVR, XGBoost, and LightGBM models provide better prediction performance after first-order derivative preprocessing. After comparing the performance of various combinations of base models applied to the first layer of a stacking ensemble model, the results showed that both the combination of XGBoost, LightGBM, and SVR models and the combination of SVR, ELM, and LightGBM models achieve the best performance. The coefficient of determination (R2) of the stacking ensemble model on the test set reaches 0.84, which improves the accuracy of the model compared with the traditional single model. The stability of the stacking ensemble model was verified by applying it to datasets of different sizes, which can replace traditional machine learning models in predicting SOC content.

Optimizing Lithium-Ion Battery Performance: Integrating Machine Learning and Explainable AI for Enhanced Energy Management

Managing the capacity of lithium-ion batteries (LiBs) accurately, particularly in large-scale applications, enhances the cost-effectiveness of energy storage systems. Less frequent replacement or maintenance of LiBs results in cost savings in the long term. Therefore, in this study, AdaBoost, gradient boosting, XGBoost, LightGBM, CatBoost, and ensemble learning models were employed to predict the discharge capacity of LiBs. The prediction performances of each model were compared based on mean absolute error (MAE), mean squared error (MSE), and R-squared values. The research findings reveal that the LightGBM model exhibited the lowest MAE (0.103) and MSE (0.019) values and the highest R-squared (0.887) value, thus demonstrating the strongest correlation in predictions. Gradient boosting and XGBoost models showed similar performance levels but ranked just below LightGBM. The competitive performance of the ensemble model indicates that combining multiple models could lead to an overall performance improvement. Furthermore, the study incorporates an analysis of key features affecting model predictions using SHAP (Shapley additive explanations) values within the framework of explainable artificial intelligence (XAI). This analysis evaluates the impact of features such as temperature, cycle index, voltage, and current on predictions, revealing a significant effect of temperature on discharge capacity. The results of this study emphasize the potential of machine learning models in LiB management within the XAI framework and demonstrate how these technologies could play a strategic role in optimizing energy storage systems.

Predictive models for delay in medical decision-making among older patients with acute ischemic stroke: a comparative study using logistic regression analysis and lightGBM algorithm.

To explore the factors affecting delayed medical decision-making in older patients with acute ischemic stroke (AIS) using logistic regression analysis and the Light Gradient Boosting Machine (LightGBM) algorithm, and compare the two predictive models. A cross-sectional study was conducted among 309 older patients aged ≥ 60 who underwent AIS. Demographic characteristics, stroke onset characteristics, previous stroke knowledge level, health literacy, and social network were recorded. These data were separately inputted into logistic regression analysis and the LightGBM algorithm to build the predictive models for delay in medical decision-making among older patients with AIS. Five parameters of Accuracy, Recall, F1 Score, AUC and Precision were compared between the two models. The medical decision-making delay rate in older patients with AIS was 74.76%. The factors affecting medical decision-making delay, identified through logistic regression and LightGBM algorithm, were as follows: stroke severity, stroke recognition, previous stroke knowledge, health literacy, social network (common factors), mode of onset (logistic regression model only), and reaction from others (LightGBM algorithm only). The LightGBM model demonstrated the more superior performance, achieving the higher AUC of 0.909. This study used advanced LightGBM algorithm to enable early identification of delay in medical decision-making groups in the older patients with AIS. The identified influencing factors can provide critical insights for the development of early prevention and intervention strategies to reduce delay in medical decisions-making among older patients with AIS and promote patients' health. The LightGBM algorithm is the optimal model for predicting the delay in medical decision-making among older patients with AIS.

#2742 Artificial intelligence accurately approximates MGPE dose response relationship in a North American Dialysis Population

Abstract Background and Aims The Anemia Control Model (ACM) is a certified medical device aimed at optimizing the management of CKD-related anemia among patients on dialysis. ACM is based on an artificial neural network (ANN) trained on over 900’000 patients treated in the European Nephrocare Network. Over the years, ACM has offered physicians hundreds of thousands of erythropoietin stimulating agent (ESA) and intravenous iron dosing suggestions and ensured high hemoglobin target achievement rate while minimizing drug use. However, evidence concerning its validity in non-European samples is still limited. Apollo Dial DB is the largest, fully anonymized, collaborative, international database of dialysis patients. The database includes detailed clinical information of more than 500,000 patients over millions of patient-months of observation with an extensive representation over geographies as well as ethnic groups. Given its detailed medical information and size, the database may offer unprecedented opportunities to develop and test equitable and fair decision support systems based on artificial intelligence. In this study we sought to re-train and validate and assess the accuracy of the ACM in an ethnically diverse sample of North American dialysis patients in Apollo Dial DB. Method We constructed the cohort for training and validation of this new ACM version by including a random sample of patients treated in the North American Fresenius Medical Care Dialysis Network. The sample accounted for 20% of the entire North American Section of the Apollo Dial DB dataset. Each patient history has been segmented in a collection of patient-months, which was the unit of analysis. The initial cohort was split in 3 datasets for training (70%), validation (20%) and test (10%). The input variables included laboratory test results, body weight, treatment data, age, sex, and previous ESA and iron prescription. The outcome of interest was the variation in serum hemoglobin concentration over 1 month of follow up. We compared the accuracy of the original ACM based on ANN (baseline model) with a newly trained algorithm based on high-performance light GBM. Results We included 44,220 patients followed by 2,277,276 patient-months overall. The mean age=60.9 ± 13.4, 59.2% were male, 56.4% were Caucasians, 32.2% African American, 3.2% Asians and 8.2% were others. At baseline, mean Hb= 10.9 ± 1.2. The original EMEA version of the Anemia Control Model (ACM) was still accurate in North American patients (mean absolute error, MAE = 0.65 g/dL). The fine-tuned light GBM model achieved a 16% reduction in error rate (MAE = 0.55 g/dL; Fig. 1). No difference in MAE was observed across sex and ethnic groups. Conclusion This study demonstrates the excellent predictive performance of the original ANN-based ACM in the North American dialysis population. Fine tuning by re-training ACM on the dataset further improved the accuracy of the model with substantial equivalence in performance across different demographics, thus ensuring algorithmic fairness. ACM is a promising tool for anemia management in ethnically diverse dialysis population and different healthcare settings.

Heavy Metal Concentration Estimation for Different Farmland Soils Based on Projection Pursuit and LightGBM with Hyperspectral Images.

Heavy metal pollution in farmland soil threatens soil environmental quality. It is an important task to quickly grasp the status of heavy metal pollution in farmland soil in a region. Hyperspectral remote sensing technology has been widely used in soil heavy metal concentration monitoring. How to improve the accuracy and reliability of its estimation model is a hot topic. This study analyzed 440 soil samples from Sihe Town and the surrounding agricultural areas in Yushu City, Jilin Province. Considering the differences between different types of soils, a local regression model of heavy metal concentrations (As and Cu) was established based on projection pursuit (PP) and light gradient boosting machine (LightGBM) algorithms. Based on the estimations, a spatial distribution map of soil heavy metals in the region was drawn. The findings of this study showed that considering the differences between different soils to construct a local regression estimation model of soil heavy metal concentration improved the estimation accuracy. Specifically, the relative percent difference (RPD) of As and Cu element estimations in black soil increased the most, by 0.30 and 0.26, respectively. The regional spatial distribution map of heavy metal concentration derived from local regression showed high spatial variability. The number of characteristic bands screened by the PP method accounted for 10-13% of the total spectral bands, effectively reducing the model complexity. Compared with the traditional machine model, the LightGBM model showed better estimation ability, and the highest determination coefficients (R2) of different soil validation sets reached 0.73 (As) and 0.75 (Cu), respectively. In this study, the constructed PP-LightGBM estimation model takes into account the differences in soil types, which effectively improves the accuracy and reliability of hyperspectral image estimation of soil heavy metal concentration and provides a reference for drawing large-scale spatial distributions of heavy metals from hyperspectral images and mastering soil environmental quality.

A personalized product recommendation model in e-commerce based on retrieval strategy

In recent years, online shopping is one of the routine parts in people’s life. It is convenient and takes less effort to purchase it. Regarding the increasing revolution of e-commerce businesses, recommendation engine plays a crucial role in them. Recommendation engines are very popular and easy to implement to their platform nowadays. Due to the extremely high competition of e-commerce businesses, the operation needs to integrate the recommender wisely. This study presents a comprehensive approach to improving user experience and engagement on e-commerce platforms through the implementation of an implicit personalized product recommendation engine. Collaborating with the H&M Group, the research combines the strength of each recommending algorithms which are collaborative filtering, popularity, and Bayesian personalized ranking to develop a robust recommendation system. By leveraging a retrieval strategy that combines multiple algorithmic techniques and evaluating candidates using machine learning models which comprise LightGBM and Deep Neural Network, the study achieves promising results. The authors utilize two popular technical metrics to evaluate their models which are mean average precision at K candidates (MAP@K) and mean average recall at K candidates (MAR@K). The empirical result indicates that the LightGBM model has remarkable performance than Deep Neural Network model, which are 0.06 versus 0.02 respectively in MAP@K and 0.03 versus 0.01 respectively in MAR@K when both recommending ways is at 50 items. Overall, this research contributes a novel framework that addresses the challenges of analyzing large-scale data, cold-start problems, and personalization, thereby enhancing the user experience, and driving sales on e-commerce platforms.

Development and application of a machine learning-based predictive model for obstructive sleep apnea screening.

To develop a robust machine learning prediction model for the automatic screening and diagnosis of obstructive sleep apnea (OSA) using five advanced algorithms, namely Extreme Gradient Boosting (XGBoost), Logistic Regression (LR), Support Vector Machine (SVM), Light Gradient Boosting Machine (LightGBM), and Random Forest (RF) to provide substantial support for early clinical diagnosis and intervention. We conducted a retrospective analysis of clinical data from 439 patients who underwent polysomnography at the Affiliated Hospital of Xuzhou Medical University between October 2019 and October 2022. Predictor variables such as demographic information [age, sex, height, weight, body mass index (BMI)], medical history, and Epworth Sleepiness Scale (ESS) were used. Univariate analysis was used to identify variables with significant differences, and the dataset was then divided into training and validation sets in a 4:1 ratio. The training set was established to predict OSA severity grading. The validation set was used to assess model performance using the area under the curve (AUC). Additionally, a separate analysis was conducted, categorizing the normal population as one group and patients with moderate-to-severe OSA as another. The same univariate analysis was applied, and the dataset was divided into training and validation sets in a 4:1 ratio. The training set was used to build a prediction model for screening moderate-to-severe OSA, while the validation set was used to verify the model's performance. Among the four groups, the LightGBM model outperformed others, with the top five feature importance rankings of ESS total score, BMI, sex, hypertension, and gastroesophageal reflux (GERD), where Age, ESS total score and BMI played the most significant roles. In the dichotomous model, RF is the best performer of the five models respectively. The top five ranked feature importance of the best-performing RF models were ESS total score, BMI, GERD, age and Dry mouth, with ESS total score and BMI being particularly pivotal. Machine learning-based prediction models for OSA disease grading and screening prove instrumental in the early identification of patients with moderate-to-severe OSA, revealing pertinent risk factors and facilitating timely interventions to counter pathological changes induced by OSA. Notably, ESS total score and BMI emerge as the most critical features for predicting OSA, emphasizing their significance in clinical assessments. The dataset will be publicly available on my Github.

Construction of diagnostic models for the progression of hepatocellular carcinoma using machine learning.

Liver cancer is one of the most prevalent forms of cancer worldwide. A significant proportion of patients with hepatocellular carcinoma (HCC) are diagnosed at advanced stages, leading to unfavorable treatment outcomes. Generally, the development of HCC occurs in distinct stages. However, the diagnostic and intervention markers for each stage remain unclear. Therefore, there is an urgent need to explore precise grading methods for HCC. Machine learning has emerged as an effective technique for studying precise tumor diagnosis. In this research, we employed random forest and LightGBM machine learning algorithms for the first time to construct diagnostic models for HCC at various stages of progression. We categorized 118 samples from GSE114564 into three groups: normal liver, precancerous lesion (including chronic hepatitis, liver cirrhosis, dysplastic nodule), and HCC (including early stage HCC and advanced HCC). The LightGBM model exhibited outstanding performance (accuracy = 0.96, precision = 0.96, recall = 0.96, F1-score = 0.95). Similarly, the random forest model also demonstrated good performance (accuracy = 0.83, precision = 0.83, recall = 0.83, F1-score = 0.83). When the progression of HCC was categorized into the most refined six stages: normal liver, chronic hepatitis, liver cirrhosis, dysplastic nodule, early stage HCC, and advanced HCC, the diagnostic model still exhibited high efficacy. Among them, the LightGBM model exhibited good performance (accuracy = 0.71, precision = 0.71, recall = 0.71, F1-score = 0.72). Also, performance of the LightGBM model was superior to that of the random forest model. Overall, we have constructed a diagnostic model for the progression of HCC and identified potential diagnostic characteristic gene for the progression of HCC.

Machine-learning classifier models for predicting sarcopenia in the elderly based on physical factors.

As the size of the elderly population gradually increases, musculoskeletal disorders, such as sarcopenia, are increasing. Diagnostic techniques such as X-rays, computed tomography, and magnetic resonance imaging are used to predict and diagnose sarcopenia, and methods using machine learning are gradually increasing. This study aimed to create a model that can predict sarcopenia using physical characteristics and activity-related variables without medical diagnostic equipment, such as imaging equipment, for the elderly aged 60 years or older. A sarcopenia prediction model was constructed using public data obtained from the Korea National Health and Nutrition Examination Survey. Models were built using Logistic Regression, Support Vector Machine (SVM), XGBoost, LightGBM, RandomForest, and Multi-layer Perceptron Neural Network (MLP) algorithms, and the feature importance of the models trained with the algorithms, except for SVM and MLP, was analyzed. The sarcopenia prediction model built with the LightGBM algorithm achieved the highest test accuracy, of 0.848. In constructing the LightGBM model, physical characteristic variables such as body mass index, weight, and waist circumference showed high importance, and activity-related variables were also used in constructing the model. The sarcopenia prediction model, which consisted of only physical characteristics and activity-related factors, showed excellent performance. This model has the potential to assist in the early detection of sarcopenia in the elderly, especially in communities with limited access to medical resources or facilities. Geriatr Gerontol Int 2024; 24: 595-602.

Machine Learning for Polymer Design to Enhance Pervaporation-Based Organic Recovery.

Pervaporation (PV) is an effective membrane separation process for organic dehydration, recovery, and upgrading. However, it is crucial to improve membrane materials beyond the current permeability-selectivity trade-off. In this research, we introduce machine learning (ML) models to identify high-potential polymers, greatly improving the efficiency and reducing cost compared to conventional trial-and-error approach. We utilized the largest PV data set to date and incorporated polymer fingerprints and features, including membrane structure, operating conditions, and solute properties. Dimensionality reduction, missing data treatment, seed randomness, and data leakage management were employed to ensure model robustness. The optimized LightGBM models achieved RMSE of 0.447 and 0.360 for separation factor and total flux, respectively (logarithmic scale). Screening approximately 1 million hypothetical polymers with ML models resulted in identifying polymers with a predicted permeation separation index >30 and synthetic accessibility score <3.7 for acetic acid extraction. This study demonstrates the promise of ML to accelerate tailored membrane designs.

Forecasting Hydrogen Vehicle Refuelling for Sustainable Transportation: A Light Gradient-Boosting Machine Model

Forecasting Hydrogen Vehicle Refuelling for Sustainable Transportation: A Light Gradient-Boosting Machine Model

Short term forecasting of base metals prices using a LightGBM and a LightGBM - ARIMA ensemble

AbstractBase metals are key materials for various industrial sectors such as electronics, construction, manufacturing, etc. Their selling price is important both for the profitability of the mining and metallurgical companies that produce and trade them, as well as for the countries whose economies rely on their exports or tax revenues as a means for national income. Prices are also critical for companies that use base metals as inputs to fabricate end products. The prediction of prices’ future movements can serve as a tool for risk mitigation and better budget planning. In this study, the logarithmic returns of base metals are forecasted using an autoregressive Light Gradient Boosting Machine (LightGBM) as well as an ensemble comprising the aforementioned algorithm and a classical time series forecasting model (i.e., ARIMA). The two models are then compared to three simpler benchmark models, namely a global mean model, an exponential smoothing model and an ARIMA model. When comparing using RMSE, the autoregressive LightGBM model outperformed the three univariate benchmark models (and the ensemble) for forecasting 6 months ahead for aluminum and nickel returns, while copper and zinc returns were forecasted better by the ensemble. Neither of the proposed models performed better than an ARIMA model when it comes to forecasting lead and tin returns.

Predicting the Occurrence of Forest Fire in the Central-South Region of China

Understanding the spatial and temporal patterns of forest fires, along with the key factors influencing their occurrence, and accurately forecasting these events are crucial for effective forest management. In the Central-South region of China, forest fires pose a significant threat to the ecological system, public safety, and economic stability. This study employs Geographic Information Systems (GISs) and the LightGBM (Light Gradient Boosting Machine) model to identify the determinants of forest fire incidents and develop a predictive model for the likelihood of forest fire occurrences, in addition to proposing a zoning strategy. The purpose of the study is to enhance our understanding of forest fire dynamics in the Central-South region of China and to provide actionable insights for mitigating the risks associated with such disasters. The findings reveal the following: (i) Spatially, fire incidents exhibit significant clustering and autocorrelation, highlighting areas with heightened likelihood. (ii) The Central-South Forest Fire Likelihood Prediction Model demonstrates high accuracy, reliability, and predictive capability, with performance metrics such as accuracy, precision, recall, and F1 scores exceeding 85% and AUC values above 89%, proving its effectiveness in forecasting the likelihood of forest fires and differentiating between fire scenarios. (iii) The likelihood of forest fires in the Central-South region of China varies across regions and seasons, with increased likelihood observed from March to May in specific provinces due to various factors, including weather conditions and leaf litter accumulation. Risks of localized fires are noted from June to August and from September to November in different areas, while certain regions continue to face heightened likelihood from December to February.

Prediction Technology of a Reservoir Development Model While Drilling Based on Machine Learning and Its Application

In order to further understand the complex spatial distribution caused by the extremely strong heterogeneity of buried hill reservoirs, this paper proposes a new method for predicting the development pattern of buried hill reservoirs based on the traditional pre-drilling prediction and post-drilling evaluation methods that mainly rely on seismic, logging, and core data, which are difficult to meet the timeliness and accuracy of drilling operations. Firstly, the box method and normalization formula are used to process and normalize the abnormal data of element logging and engineering logging, and then the stepwise regression analysis method is used to optimize the sensitive parameters of element logging and engineering logging. The Light Gradient Boosting Machine (LightGBM) algorithm, deep neural network (DNN), and support vector machine (SVM) are used to establish a new method for predicting the development pattern of buried hill reservoirs. Lastly, a comprehensive evaluation index F1 score for the model is established to evaluate the prediction model for the development pattern of buried hill reservoirs. The F1 score value obtained from this model’s comprehensive evaluation index indicates that the LightGBM model achieves the highest accuracy, with 96.7% accuracy in identifying weathered zones and 95.8% accuracy in identifying interior zones. The practical application demonstrates that this method can rapidly and accurately predict the development mode of buried hill reservoirs while providing a new approach for efficient on-site exploration and decision-making in oil and gas field developments. Consequently, it effectively promotes exploration activities as well as enhances the overall process of oil and gas reservoir exploration.

Development and Validation of a Prediction Model for Acute Hypotensive Events in Intensive Care Unit Patients.

Background: Persistent hypotension in the intensive care unit (ICU) is associated with increased mortality. Predicting acute hypotensive events can lead to timely intervention. We aimed to develop a prediction model of acute hypotensive events in patients admitted to the ICU. Methods: We included adult patients admitted to the Nagoya City University (NCU) Hospital ICU between January 2018 and December 2021 for model training and internal validation. The MIMIC-III database was used for external validation. A hypotensive event was defined as a mean arterial pressure < 60 mmHg for at least 5 min in 10 min. The input features were age, sex, and time-series data for vital signs. We compared the area under the receiver-operating characteristic curve (AUROC) of three machine-learning algorithms: logistic regression, the light gradient boosting machine (LightGBM), and long short-term memory (LSTM). Results: Acute hypotensive events were found in 1325/1777 (74.6%) and 2691/5266 (51.1%) of admissions in the NCU and MIMIC-III cohorts, respectively. In the internal validation, the LightGBM model had the highest AUROC (0.835), followed by the LSTM (AUROC 0.834) and logistic regression (AUROC 0.821) models. Applying only blood pressure-related features, the LSTM model achieved the highest AUROC (0.843) and consistently showed similar results in external and internal validation. Conclusions: The LSTM model using only blood pressure-related features had the highest AUROC with comparable performance in external validation.

LightGBM Model for Detecting Fraud in Online Financial Transactions

With the rapid development of online finance, combined with its significant potential threats, nowadays it has increasingly drawn attention to the issue of fraud in online payment transactions. Due to the large and diverse nature of online payment data and the binary nature of fraud detection, this study implements the machine learning algorithm LightGBM to detect fraudulent activities. Through data exploration and data preprocessing, the data is refined. Leveraging violin plots to observe the distribution of variables in fraudulent and non-fraudulent activities, and then an insignificant variable is removed. Next, the processed data is fed into the LightGBM model. The result yields a fraud prediction accuracy of 99.5%, shows a very high precision. Additionally, the study verifies the algorithm's robustness and generalization ability. Overall, this research holds significant importance, given its findings and implications for addressing fraud in the rapidly evolving landscape of online financial transactions.

Development and validation of machine learning models to predict unplanned hospitalizations of patients with diabetes within the next 12 months

BACKGROUND: The incidence of diabetes mellitus (DM) both in the Russian Federation and in the world has been steadily increasing for several decades. Stable population growth and current epidemiological characteristics of DM lead to enormous economic costs and significant social losses throughout the world. The disease often progresses with the development of specific complications, while significantly increasing the likelihood of hospitalization. The creation and inference of a machine learning model for predicting hospitalizations of patients with DM to an inpatient medical facility will make it possible to personalize the provision of medical care and optimize the load on the entire healthcare system.AIM: Development and validation of models for predicting unplanned hospitalizations of patients with diabetes due to the disease itself and its complications using machine learning algorithms and data from real clinical practice.MATERIALS AND METHODS: 170,141 depersonalized electronic health records of 23,742 diabetic patients were included in the study. Anamnestic, constitutional, clinical, instrumental and laboratory data, widely used in routine medical practice, were considered as potential predictors, a total of 33 signs. Logistic regression (LR), gradient boosting methods (LightGBM, XGBoost, CatBoost), decision tree-based methods (RandomForest and ExtraTrees), and a neural network-based algorithm (Multi-layer Perceptron) were compared. External validation was performed on the data of the separate region of Russian Federation.RESULTS: The best results and stability to external validation data were shown by the LightGBM model with an AUC of 0.818 (95% CI 0.802–0.834) in internal testing and 0.802 (95% CI 0.773–0.832) in external validation.CONCLUSION: The metrics of the best model were superior to previously published studies. The results of external validation showed the relative stability of the model to new data from another region, that reflects the possibility of the model’s application in real clinical practice.

Estimating the Thumb Rotation Angle by Using a Tablet Device With a Posture Estimation Artificial Intelligence Model.

MediaPipe Hand (MediaPipe) is an artificial intelligence (AI)-based pose estimation library. In this study, MediaPipe was combined with four machine learning (ML) models to estimate the rotation angle of the thumb. Videos of the right hands of 15 healthy volunteers were recorded and processed into 9000 images. The rotation angle of the thumb (defined as angle θ from the palmar plane, which is defined as 0°) was measured using an angle measuring device, expressed in a radiansystem. Angle θ was then estimated by the ML model by using parameters calculated from the hand coordinates detected by MediaPipe. The linear regression model showed a root mean square error (RMSE) of 12.23, a mean absolute error (MAE) of 9.9, and a correlation coefficient of 0.91. The ElasticNet model showed an RMSE of 12.23, an MAE of 9.95, and a correlation coefficient of 0.91; the support vector machine (SVM) model showed an RMSE of 4.7, an MAE of 2.5, and a correlation coefficient of 0.99. The LightGBM model achieved high values: an RMSE of 4.58, an MAE of 2.62, and a correlation coefficient of 0.99. Based on these findings, we concluded that the thumb rotation angle can be estimated with high accuracy by combining MediaPipe and ML.