Gradient Tree Boosting Research Articles

Predicting distant metastasis in nasopharyngeal carcinoma using gradient boosting tree model based on detailed magnetic resonance imaging reports.

Development of distant metastasis (DM) is a major concern during treatment of nasopharyngeal carcinoma (NPC). However, studies have demonstrated improved distant control and survival in patients with advanced NPC with the addition of chemotherapy to concomitant chemoradiotherapy. Therefore, precise prediction of metastasis in patients with NPC is crucial. To develop a predictive model for metastasis in NPC using detailed magnetic resonance imaging (MRI) reports. This retrospective study included 792 patients with non-distant metastatic NPC. A total of 469 imaging variables were obtained from detailed MRI reports. Data were stratified and randomly split into training (50%) and testing sets. Gradient boosting tree (GBT) models were built and used to select variables for predicting DM. A full model comprising all variables and a reduced model with the top-five variables were built. Model performance was assessed by area under the curve (AUC). Among the 792 patients, 94 developed DM during follow-up. The number of metastatic cervical nodes (30.9%), tumor invasion in the posterior half of the nasal cavity (9.7%), two sides of the pharyngeal recess (6.2%), tubal torus (3.3%), and single side of the parapharyngeal space (2.7%) were the top-five contributors for predicting DM, based on their relative importance in GBT models. The testing AUC of the full model was 0.75 (95% confidence interval [CI]: 0.69-0.82). The testing AUC of the reduced model was 0.75 (95%CI: 0.68-0.82). For the whole dataset, the full (AUC = 0.76, 95%CI: 0.72-0.82) and reduced models (AUC = 0.76, 95%CI: 0.71-0.81) outperformed the tumor node-staging system (AUC = 0.67, 95%CI: 0.61-0.73). The GBT model outperformed the tumor node-staging system in predicting metastasis in NPC. The number of metastatic cervical nodes was identified as the principal contributing variable.

Machine learning analysis of thermophysical and thermohydraulic properties in ethylene glycol- and glycerol-based SiO2 nanofluids

The study investigates the heat transfer and friction factor properties of ethylene glycol and glycerol-based silicon dioxide nanofluids flowing in a circular tube under continuous heat flux circumstances. This study tackles the important requirement for effective thermal management in areas such as electronics cooling, the automobile industry, and renewable energy systems. Previous research has encountered difficulties in enhancing thermal performance while handling the increased friction factor associated with nanofluids. This study conducted experiments in the Reynolds number range of 1300 to 21,000 with particle volume concentrations of up to 1.0%. Nanofluids exhibited superior heat transfer coefficients and friction factor values than the base liquid values. The highest enhancement in heat transfer was 5.4% and 8.3% for glycerol and ethylene glycol -based silicon dioxide Nanofluid with a relative friction factor penalty of ∼30% and 75%, respectively. To model and predict the complicated, nonlinear experimental data, five machine learning approaches were used: linear regression, random forest, extreme gradient boosting, adaptive boosting, and decision tree. Among them, the decision tree-based model performed well with few errors, while the random forest and extreme gradient boosting models were also highly accurate. The findings indicate that these advanced machine learning models can accurately anticipate the thermal performance of nanofluids, providing a dependable tool for improving their use in a variety of thermal systems. This study's findings help to design more effective cooling solutions and improve the sustainability of energy systems.

Geometric data analysis-based machine learning for two-dimensional perovskite design

With extraordinarily high efficiency, low cost, and excellent stability, 2D perovskite has demonstrated a great potential to revolutionize photovoltaics technology. However, inefficient material structure representations have significantly hindered artificial intelligence (AI)-based perovskite design and discovery. Here we propose geometric data analysis (GDA)-based perovskite structure representation and featurization and combine them with learning models for 2D perovskite design. Both geometric properties and periodicity information of the material unit cell, are fully characterized by a series of 1D functions, i.e., density fingerprints (DFs), which are mathematically guaranteed to be invariant under different unit cell representations and stable to structure perturbations. Element-specific DFs, which are based on different site combinations and atom types, are combined with gradient boosting tree (GBT) model. It has been found that our GDA-based learning models can outperform all existing models, as far as we know, on the widely used new materials for solar energetics (NMSE) databank.

Advancing Skin Cancer Prediction Using Ensemble Models

There are many different kinds of skin cancer, and an early and precise diagnosis is crucial because skin cancer is both frequent and deadly. The key to effective treatment is accurately classifying the various skin cancers, which have unique traits. Dermoscopy and other advanced imaging techniques have enhanced early detection by providing detailed images of lesions. However, accurately interpreting these images to distinguish between benign and malignant tumors remains a difficult task. Improved predictive modeling techniques are necessary due to the frequent occurrence of erroneous and inconsistent outcomes in the present diagnostic processes. Machine learning (ML) models have become essential in the field of dermatology for the automated identification and categorization of skin cancer lesions using image data. The aim of this work is to develop improved skin cancer predictions by using ensemble models, which combine numerous machine learning approaches to maximize their combined strengths and reduce their individual shortcomings. This paper proposes a fresh and special approach for ensemble model optimization for skin cancer classification: the Max Voting method. We trained and assessed five different ensemble models using the ISIC 2018 and HAM10000 datasets: AdaBoost, CatBoost, Random Forest, Gradient Boosting, and Extra Trees. Their combined predictions enhance the overall performance with the Max Voting method. Moreover, the ensemble models were fed with feature vectors that were optimally generated from the image data by a genetic algorithm (GA). We show that, with an accuracy of 95.80%, the Max Voting approach significantly improves the predictive performance when compared to the five ensemble models individually. Obtaining the best results for F1-measure, recall, and precision, the Max Voting method turned out to be the most dependable and robust. The novel aspect of this work is that skin cancer lesions are more robustly and reliably classified using the Max Voting technique. Several pre-trained machine learning models’ benefits are combined in this approach.

Understanding the role of driver behaviors and performance in safety-critical events: Application of machine learning

Human factors contribute in some way to almost 93% of road crashes. Because the unavailability of good pre-crash data, the contribution of human factors to safety-critical events (SCEs) and the prediction of crashes using real-world data is lightly researched. This study provides predictive accuracy by harnessing unique real-world naturalistic driving study (NDS) data, which includes dynamic pre-crash information about driving behavior and performance. After cleaning and preprocessing, a final subsample (N = 9,237) was used and split into training and test samples. For consistent comparison of variables’ importance in statistical and machine learning (ML) models, the dominance analysis uncovered the most important predictors used by the ordered Probit model. Next, three non-parametric supervised ML methods, because promising prediction performance and cost-effectiveness, including Naïve Bayes, K-Nearest Neighbors, and Gradient Boosting Tree (GBT) were used. The overall out-of-sample prediction accuracy for the ordered Probit model was 85.75% which was lower than all three ML methods. The GBT showed the highest (91.23%) out-of-sample prediction accuracy. The availability of pre-crash naturalistic data helps significantly improve the prediction accuracy of SCEs as cumulative importance for all available human factors in the GBT classifier was 94%. For practical applications, refer to the article.

A GA-stacking ensemble approach for forecasting energy consumption in a smart household: A comparative study of ensemble methods

The considerable amount of energy utilized by buildings has led to various environmental challenges that adversely impact human existence. Predicting buildings' energy usage is commonly acknowledged as encouraging energy efficiency and enabling well-informed decision-making, ultimately leading to decreased energy consumption. Implementing eco-friendly architectural designs is paramount in mitigating energy consumption, particularly in recently constructed structures. This study utilizes clustering analysis on the original dataset to capture complex consumption patterns over various periods. The analysis yields two distinct subsets that represent low and high consumption patterns and an additional subset that exclusively encompasses weekends, attributed to the specific behavior of occupants. Ensemble models have become increasingly popular due to advancements in machine learning techniques. This research utilizes three discrete algorithms, namely Artificial Neural Network (ANN), K-nearest neighbors (KNN), and Decision Trees (DT). In addition, the application employs three more machine learning algorithms bagging and boosting: Random Forest (RF), Extreme Gradient Boosting (XGB), and Gradient Boosting Trees (GBT). To augment the accuracy of predictions, a stacking ensemble methodology is employed, wherein the forecasts generated by many algorithms are combined. Given the obtained outcomes, a thorough examination is undertaken, encompassing the techniques of stacking, bagging, and boosting, to conduct a comprehensive comparative study. It is pertinent to highlight that the stacking technique consistently exhibits superior performance relative to alternative ensemble methodologies across a spectrum of heterogeneous datasets. Furthermore, using a genetic algorithm enables the optimization of the combination of base learners, resulting in a notable enhancement in prediction accuracy. After implementing this optimization technique, GA-Stacking demonstrated remarkable performance in Mean Absolute Percentage Error (MAPE) scores. The improvement observed was substantial, surpassing 90 percent for all datasets. In addition, in subset-1, subset-2, and subset-3, the achieved R2 scores were 0.983, 0.985, and 0.999, respectively. This represents a substantial advancement in forecasting the energy consumption of residential buildings. Such progress underscores the potential advantages of integrating this framework into the practices of building designers, thereby fostering informed decision-making, design management, and optimization prior to construction.

Estimation οf Wave Overtopping Discharges at Coastal Structures with Combined Slopes using Machine Learning Techniques

Coastal defense structures are of paramount importance in protecting coastal communities from the adverse impacts of severe weather events and flooding. This study uses machine learning techniques, specifically Decision Tree (DT), Gradient Boosted Tree (GBT), and Support Vector Machine (SVM) models, to estimate wave overtopping discharge at coastal structures with combined slopes employing the recently built EurOtop database. The models were evaluated by deploying statistical metrics and Taylor diagram visualization. The GBT model demonstrated a high level of accuracy in predicting wave-overtopping discharge. Compared to the other models, the scatter index of GBT (0.392) was lower than that of DT (0.512) and SVM (0.823). In terms of the R-index, GBT (0.991) was superior to DT (0.977) and SVM (0.943). The GBT results were also compared with those of previous works. The findings showed that the GBT model significantly decreased the overall error and provided accurate estimations of the wave-overtopping discharge.

Photon classification with Gradient Boosted Trees at CLAS12

Dihadron semi-inclusive deep inelastic scattering (SIDIS) of 10.6 GeV longitudinally polarized electrons off the proton has been measured using the CLAS12 detector at Jefferson Lab. Two separate channels, π + π 0 and π - π 0, were analyzed, requiring the reconstruction of diphoton pairs. In this analysis, we addressed the problem of false neutral particles being reconstructed by CLAS12's event builder, polluting the otherwise physical combinatorial background underneath the π 0 peak. A photon classifier using a Gradient Boosted Trees (GBTs) architecture was trained with Monte Carlo simulations to reduce the amount of background π 0's. We show that the nearest-neighbor features learned by the model lead to a substantial increase in signal vs. background discrimination compared to previous CLAS12 π^0 analyses. The machine learning approach recovers several times more dihadron statistics for the dataset.

Machine learning approaches for predicting postpartum hemorrhage: a comprehensive systematic literature review

Postpartum hemorrhage (PPH) represents a significant threat to maternal health, particularly in developing countries, where it remains a leading cause of maternal mortality. Unfortunately, only 60% of pregnant women at high risk for PPH are identified, leaving 40% undetected until they experience PPH. To address this critical issue and ensure timely intervention, leveraging rapidly advancing technology with machine learning (ML) methodologies for maternal health prediction is imperative. This review synthesizes findings from 43 selected research articles, highlighting the predominant ML techniques employed in PPH prediction. Among these, logistic regression (LR), extreme gradient boosting (XGB), random forest (RF), and decision tree (DT) emerge as the most frequently utilized methods. By harnessing the power of ML, we aim to foster technological advancements in the healthcare sector, with a particular focus on maternal health and ultimately contribute to the reduction of maternal mortality rates worldwide.

Deep Neural Networks and Tabular Data: A Survey.

Heterogeneous tabular data are the most commonly used form of data and are essential for numerous critical and computationally demanding applications. On homogeneous datasets, deep neural networks have repeatedly shown excellent performance and have therefore been widely adopted. However, their adaptation to tabular data for inference or data generation tasks remains highly challenging. To facilitate further progress in the field, this work provides an overview of state-of-the-art deep learning methods for tabular data. We categorize these methods into three groups: data transformations, specialized architectures, and regularization models. For each of these groups, our work offers a comprehensive overview of the main approaches. Moreover, we discuss deep learning approaches for generating tabular data and also provide an overview over strategies for explaining deep models on tabular data. Thus, our first contribution is to address the main research streams and existing methodologies in the mentioned areas while highlighting relevant challenges and open research questions. Our second contribution is to provide an empirical comparison of traditional machine learning methods with 11 deep learning approaches across five popular real-world tabular datasets of different sizes and with different learning objectives. Our results, which we have made publicly available as competitive benchmarks, indicate that algorithms based on gradient-boosted tree ensembles still mostly outperform deep learning models on supervised learning tasks, suggesting that the research progress on competitive deep learning models for tabular data is stagnating. To the best of our knowledge, this is the first in-depth overview of deep learning approaches for tabular data; as such, this work can serve as a valuable starting point to guide researchers and practitioners interested in deep learning with tabular data.

Prediction of viral families and hosts of single-stranded RNA viruses based on K-Mer coding from phylogenetic gene sequences

There are billions of virus species worldwide, and viruses, the smallest parasitic entities, pose a serious threat. Therefore, fighting associated disorders requires an understanding of the genetic structure of viruses. Considering the wide diversity and rapid evolution of viruses, there is a critical need to quickly and accurately classify viral species and their potential hosts to better understand transmission dynamics, facilitating the development of targeted therapies. Recognizing this, this study has investigated the classes of RNA viruses based on their genomic sequences using Machine Learning (ML) and Deep Learning (DL) models. The PhyVirus dataset, consisting of pathogenic Single-stranded RNA viruses of Baltimore group four (+ssRNA) and five (-ssRNA) with different hosts and species, was analyzed. The dataset containing viral gene sequences was analyzed using the K-Mer coding technique, which is based on base words of various lengths. The study used classical ML algorithms (Random Forest, Gradient Boosting and Extra Trees) and the Fully Connected Deep Neural Network, a Deep Learning algorithm, to predict viral families and hosts. Detailed analyses were performed on the classifier performance in scenarios with different train-test ratios and different word lengths (k-values) for K-Mer. The observed results show that Fully Connected Deep Neural Network has a high success rate of 99.60 % in predicting virus families. In predicting virus hosts, the Extra Trees classifier achieved the highest success rate of 81.53 %. This study is considered to be the first classification study in the literature on this dataset, which has a very large family and host diversity consisting of gene sequences of Single-stranded RNA viruses. Our detailed investigations on how varying word lengths based on K-Mer coding in gene sequences affect the classification into viral families and hosts make this study particularly valuable. This study shows that ML and DL methods have the potential to produce valuable results in phylogenetic studies. In addition, the results and high-performance values show that these methods can be successfully used in regenerative applications of gene sequences or in studies such as the elimination of losses in gene sequences.

Political social media bot detection: Unveiling cutting-edge feature selection and engineering strategies in machine learning model development

Over time, social media bots (SMBs), specifically political SMBs, have played a crucial role in influencing and spreading misinformation, manipulating public opinion, and harassing and intimidating users of online social networks (OSNs). This article aims to study previous works on the detection and analysis of political SMB activities and address critical challenges that significantly impact the effectiveness of SMB detection models. These challenges include feature engineering, feature selection (FS), and model implementation. Over 33 features were extracted from the Twibot-20 dataset, including content, user information, network, behavior, and temporal features. Various FS techniques are explored and compared to select the optimal features, comprising basic, filter, wrapper, embedded, and hybrid. The optimal features are then employed to train multiple machine-learning algorithms. To balance the dataset, the synthetic minority oversampling technique coupled with edited nearest neighbors (Smote-ENN) is used. The results showed an improvement in model performance, from an initial Area Under the Curve (AUC) of 90.40 % and accuracy of 81.60 % using the original set to a score of 99.50 % for the test set and 100 % for the training set in all used metrics. Decision Trees, Random Forest, Gradient Boosting, Adaboost, XGB, and Extra Trees emerge as the most effective for detecting political SMBs.

Exploring the Differences in Tree Species Classification between Typical Forest Regions in Northern and Southern China

Focusing on the trend of continuously seeking high-precision tree species classification results in small areas from the perspectives of sensors and classification algorithms. This study aimed to explore the effects of data sources, classifiers, and seasons on classification accuracy in regions with significant environmental variation, examining patterns of tree species classification to enhance the transferability of classification. Considering two typical forest distribution regions in the north and south of China, this study utilized the revisitation cycle and open-source advantages of Sentinel-2 and Landsat-8. Leveraging the Google Earth Engine (GEE) platform, this study captured spectral features, vegetation indices, and texture features for single seasonal and seasonal combination images. With the assistance of Sentinel-1A and SRTM (Shuttle Radar Topography Mission) DEM (Digital Elevation Model), backscattering coefficient features and topographical features were extracted and input with features captured from Sentinel-2 and Landsat-8 into three types of classifiers: random forest (RF), support vector machine (SVM), and gradient tree boosting (GTB) for major tree species classification. In this research, we discovered that the best classification for single season in the northern study area was spring, whereas, for the southern study area, it was winter. Seasonal combination images effectively improved the classification accuracy of single seasonal images, with Sentinel-2 imagery displaying better classification performance compared to Landsat-8, and the optimal classifier differing between the north and the south. The inclusion of topographical or backscattering coefficient features in the four-season combination imagery contributed to improvements in classification accuracy, with topographical features significantly enhancing the classification performance in the topographically varied southern study area. The evaluation of feature importance indicated that elevation was the most critical feature for classification, while spectral features and vegetation indices were also significant. In the southern study area with large topographical discrepancies, subdividing into different terrain units led to improved tree species classification accuracy in medium-altitude, gentle slope areas. These findings provide insights into the regularity of enhancing tree species classification accuracy in environmentally diverse areas through the use of multi-source remote sensing data and multi-seasonal imagery. Consequently, the results offer a reference for the identification of tree species across large areas and the creation of spatial distribution maps.

Employee Attrition Prediction in the USA: A Machine Learning Approach for HR Analytics and Talent Retention Strategies

In the dynamic business domain in the USA, human capital is one of the most instrumental assets for companies. Maintaining high performance and reducing employee attrition has become an utmost priority in the USA since the costs related to employee attrition can be significant. The chief objective of this study was to explore the application of machine learning in terms of forecasting employee attrition and its ramifications for HR analytics and talent retention strategies. In this study, the investigator used Jupyter Notebook, an interactive platform for Python users, to design machine learning algorithms. The dataset utilized in this research was attained from the IBM Human Resource workforce attrition survey dataset. In the current research, the investigator proposed an array of machine learning models, most notably, Decision Tree, Ada-boost classifier, Random Forest, and gradient-boosted classifier. By referring to the model’s performance evaluation, it was apparent that the Random Forest algorithm had the highest accuracy, followed by Gradient Boosting and Decision Tree respectively. AdaBoost had the lowest accuracy. Concerning precision, the Random Forest algorithm again had the highest precision followed by Gradient Boosting and AdaBoost accordingly. By implementing the proposed models’ organizations in the USA can identify high-performing employees at risk of quitting, and subsequently take proactive steps to retain them, saving significant organizational resources. Ultimately, the proposed machine learning techniques can assist the government in maintaining high-performing employees, reducing the impact of labor shortages, and maintaining business continuity.

Application of machine-learning model to optimize colonic adenoma detection in India.

There is limited data on the prevalence and risk factors of colonic adenoma from the Indian sub-continent. We aimed at developing a machine-learning model to optimize colonic adenoma detection in a prospective cohort. All consecutive adult patients undergoing diagnostic colonoscopy were enrolled between October 2020 and November 2022. Patients with a high risk of colonic adenoma were excluded. The predictive model was developed using the gradient-boosting machine (GBM)-learning method. The GBM model was optimized further by adjusting the learning rate and the number of trees and 10-fold cross-validation. Total 10,320 patients (mean age 45.18 ± 14.82years; 69% men) were included in the study. In the overall population, 1152 (11.2%) patients had at least one adenoma. In patients with age > 50years, hospital-based adenoma prevalence was 19.5% (808/4144). The area under the receiver operating curve (AUC) (SD) of the logistic regression model was 72.55% (4.91), while the AUCs for deep learning, decision tree, random forest and gradient-boosted tree model were 76.25% (4.22%), 65.95% (4.01%), 79.38% (4.91%) and 84.76% (2.86%), respectively. After model optimization and cross-validation, the AUC of the gradient-boosted tree model has increased to 92.2% (1.1%). Machine-learning models may predict colorectal adenoma more accurately than logistic regression. A machine-learning model may help optimize the use of colonoscopy to prevent colorectal cancers. ClinicalTrials.gov (ID: NCT04512729).

Detecting and mitigating security anomalies in Software-Defined Networking (SDN) using Gradient-Boosted Trees and Floodlight Controller characteristics

Cutting-edge and innovative software solutions are provided to address network security, network virtualization, and other network-related challenges in highly congested SDN-powered networks. However, these networks are susceptible to the same security issues as traditional networks. For instance, SDNs are significantly vulnerable to distributed denial of service (DDoS) attacks. Previous studies have suggested various anomaly detection techniques based on machine learning, statistical analysis, or entropy measurement to combat DDoS attacks and other security threats in SDN networks. However, these techniques face challenges such as collecting sufficient and relevant flow data, extracting and selecting the most informative features, and choosing the best model for identifying and preventing anomalies. This paper introduces a new and advanced multi-stage modular approach for anomaly detection and mitigation in SDN networks. The approach consists of four modules: data collection, feature selection, anomaly classification, and anomaly response. The approach utilizes the NetFlow standard to gather data and generate a dataset, employs the Information Gain Ratio (IGR) to select the most valuable features, uses gradient-boosted trees (GBT), and leverages Representational State Transfer Application Programming Interfaces (REST API) and Static Entry Pusher within the floodlight controller to construct an exceptionally efficient structure for detecting and mitigating anomalies in SDN design. We conducted experiments on a synthetic dataset containing 15 types of anomalies, such as DDoS attacks, port scans, worms, etc. We compared our model with four existing techniques: SVM, KNN, DT, and RF. Experimental results demonstrate that our model outperforms the existing techniques in terms of enhancing Accuracy (AC) and Detection Rate (DR) while simultaneously reducing Classification Error (CE) and False Alarm Rate (FAR) to 98.80 %, 97.44 %, 1.2 %, and 0.38 %, respectively.

Predicting malaria outbreak in The Gambia using machine learning techniques.

Malaria is the most common cause of death among the parasitic diseases. Malaria continues to pose a growing threat to the public health and economic growth of nations in the tropical and subtropical parts of the world. This study aims to address this challenge by developing a predictive model for malaria outbreaks in each district of The Gambia, leveraging historical meteorological data. To achieve this objective, we employ and compare the performance of eight machine learning algorithms, including C5.0 decision trees, artificial neural networks, k-nearest neighbors, support vector machines with linear and radial kernels, logistic regression, extreme gradient boosting, and random forests. The models are evaluated using 10-fold cross-validation during the training phase, repeated five times to ensure robust validation. Our findings reveal that extreme gradient boosting and decision trees exhibit the highest prediction accuracy on the testing set, achieving 93.3% accuracy, followed closely by random forests with 91.5% accuracy. In contrast, the support vector machine with a linear kernel performs less favorably, showing a prediction accuracy of 84.8% and underperforming in specificity analysis. Notably, the integration of both climatic and non-climatic features proves to be a crucial factor in accurately predicting malaria outbreaks in The Gambia.

Modeling driver injury severity using machine learning algorithms

This study planned to predict and analyze the driver injury severity (DIS) using 12 machine learning (ML) algorithms. Police reports of single- and two-vehicle accidents that occurred during 2011–2020 in the two cities of India (Itanagar and Imphal) were used in this study. The best-performing model to predict the DIS for Itanagar was Gradient Boosting Trees (GBT). “Causes of Accident” variable had shown maximum impact on the DIS. In the case of Imphal, it was the GBT, Extra Trees, and Random Forest models across all k-fold cross-validation for train ratios 0.70, 0.80, and 0.90, respectively. “Causes of Accident” and “Vehicle Type” had shown maximum impact on the DIS. These results reveal that the ML models can be applied in hilly areas to predict and identify the important factors that affect DIS. Transportation authorities can analyze road accident data using these models while implementing various road safety measures.

Using Transfer Learning and XGBoost for Early Detection of Fires in Offshore Wind Turbine Units

To improve the power generation efficiency of offshore wind turbines and address the problem of high fire monitoring and warning costs, we propose a data-driven fire warning method based on transfer learning for wind turbines in this paper. This paper processes wind turbine operation data in a SCADA system. It uses an extreme gradient-boosting tree (XGBoost) algorithm to build an offshore wind turbine unit fire warning model with a multiparameter prediction function. This paper selects some parameters from the dataset as input variables for the model, with average cabin temperature, average outdoor temperature, average cabin humidity, and average atmospheric humidity as output variables. This paper analyzes the distribution information of input and output variables and their correlation, analyzes the predicted difference, and then provides an early warning for wind turbine fires. This paper uses this fire warning model to transfer learning to different models of offshore wind turbines in the same wind farm to achieve fire warning. The experimental results show that the prediction performance of the multiparameter is accurate, with an average MAPE of 0.016 and an average RMSE of 0.795. It is better than the average MAPE (0.051) and the average RMSE (2.020) of the prediction performance of a backpropagation (BP) neural network, as well as the average MAPE (0.030) and the average RMSE (1.301) of the prediction performance of random forest. The transfer learning model has good prediction performance, with an average MAPE of 0.022 and an average RMSE of 1.469.

Predictive Modelling of Postpartum Haemorrhage Using Early Risk Factors: A Comparative Analysis of Statistical and Machine Learning Models.

Postpartum haemorrhage (PPH) is a significant cause of maternal morbidity and mortality worldwide, particularly in low-resource settings. This study aimed to develop a predictive model for PPH using early risk factors and rank their importance in terms of predictive ability. The dataset was obtained from an observational case-control study in northern Rwanda. Various statistical models and machine learning techniques were evaluated, including logistic regression, logistic regression with elastic-net regularisation, Random Forests, Extremely Randomised Trees, and gradient-boosted trees with XGBoost. The Random Forest model, with an average sensitivity of 80.7%, specificity of 71.3%, and a misclassification rate of 12.19%, outperformed the other models, demonstrating its potential as a reliable tool for predicting PPH. The important predictors identified in this study were haemoglobin level during labour and maternal age. However, there were differences in PPH risk factor importance in different data partitions, highlighting the need for further investigation. These findings contribute to understanding PPH risk factors, highlight the importance of considering different data partitions and implementing cross-validation in predictive modelling, and emphasise the value of identifying the appropriate prediction model for the application. Effective PPH prediction models are essential for improving maternal health outcomes on a global scale. This study provides valuable insights for healthcare providers to develop predictive models for PPH to identify high-risk women and implement targeted interventions.