Stratified 10-fold Cross-validation Research Articles

Implementation of Information Gain Ratio and Particle Swarm Optimization in the Sentiment Analysis Classification of Covid-19 Vaccine Using Support Vector Machine

In the current digital era, sentiment analysis has become an effective method for identifying and interpreting public opinions on various topics, including public health issues such as COVID-19 vaccination. Vaccination is a crucial measure in tackling this pandemic, but there are still a number of people who are skeptical and reluctant to receive the COVID-19 vaccine. This public perception is largely influenced by, including information received from social media and online platforms. Therefore, sentiment analysis of the COVID-19 vaccine is one way to understand the public's perception of the COVID-19 vaccine. This research has the purpose to enhance the classification performance in sentiment analysis of COVID-19 vaccines by implementing Information Gain Ratio (IGR) and Particle Swarm Optimization (PSO) on the Support Vector Machine (SVM). With a dataset of 2000 entries consisting of 1000 positive labels and 1000 negative labels, validation was performed through a combination of data splitting with an 80:20 ratio and stratified 10-Fold cross-validation. Applying the basic SVM, an accuracy of 0.794 and an AUC value of 0.890 were obtained. Integration with Information Gain Ratio (IGR) feature selection improved the accuracy to 0.814 and an AUC of 0.907. Furthermore, through the combination of SVM based on PSO and IGR, the accuracy significantly improved to 0.837 with an AUC of 0.913. These results demonstrate that the combination of feature selection techniques and parameter optimization can enhance the performance of sentiment classification towards COVID-19 vaccines. The conclusions drawn from this research indicate that the integration of IGR and PSO positively contributes to the effectiveness and predictive capability of the SVM model in sentiment classification tasks.

Distributional Representation of Cyclic Alternating Patterns for A-Phase Classification in Sleep EEG

This article describes a detailed methodology for the A-phase classification of the cyclic alternating patterns (CAPs) present in sleep electroencephalography (EEG). CAPs are a valuable EEG marker of sleep instability and represent an important pattern with which to analyze additional characteristics of sleep processes, and A-phase manifestations have been linked to some specific conditions. CAP phase detection and classification are not commonly carried out routinely due to the time and attention this problem requires (and if present, CAP labels are user-dependent, visually evaluated, and hand-made); thus, an automatic tool to solve the CAP classification problem is presented. The classification experiments were carried out using a distributional representation of the EEG data obtained from the CAP Sleep Database. For this purpose, data symbolization was performed using the one-dimensional symbolic aggregate approximation (1d-SAX), followed by the vectorization of symbolic data with a trained Doc2Vec model and a final classification with ten classic machine learning models for two separate validation strategies. The best results were obtained using a support vector classifier with a radial basis kernel. For hold-out validation, the best F1 Score was 0.7651; for stratified 10-fold cross-validation, the best F1 Score was 0.7611 ± 0.0133. This illustrates that the proposed methodology is suitable for CAP classification.

Parameter-Efficient Densely Connected Dual Attention Network for Phonocardiogram Classification.

Cardiac auscultation, exhibited by phonocardiogram (PCG), is a non-invasive and low-cost diagnostic method for cardiovascular diseases (CVDs). However, deploying it in practice is quite challenging, due to the inherent murmurs and a limited number of supervised samples in heart sound data. To solve these problems, not only heart sound analysis based on handcrafted features, but also computer-aided heart sound analysis based on deep learning have been extensively studied in recent years. Though with elaborate design, most of these methods still use additional pre-processing to improve classification performance, which heavily relies on time-consuming experienced engineering. In this paper, we propose a parameter-efficient densely connected dual attention network (DDA) for heart sound classification. It combines two advantages simultaneously of the purely end-to-end architecture and enriched contextual representations of the self-attention mechanism. Specifically, the densely connected structure can automatically extract the information flow of heart sound features hierarchically. Alongside, improving contextual modeling capabilities, the dual attention mechanism adaptively aggregates local features with global dependencies via a self-attention mechanism, which captures the semantic interdependencies across position and channel axes respectively. Extensive experiments across stratified 10-fold cross-validation strongly evidence that our proposed DDA model surpasses current 1D deep models on the challenging Cinc2016 benchmark with significant computational efficiency.

Machine learning pipeline for blood culture outcome prediction using Sysmex XN-2000 blood sample results in Western Australia

BackgroundBloodstream infections (BSIs) are a significant burden on the global population and represent a key area of focus in the hospital environment. Blood culture (BC) testing is the standard diagnostic test utilised to confirm the presence of a BSI. However, current BC testing practices result in low positive yields and overuse of the diagnostic test. Diagnostic stewardship research regarding BC testing is increasing, and becoming more important to reduce unnecessary resource expenditure and antimicrobial use, especially as antimicrobial resistance continues to rise. This study aims to establish a machine learning (ML) pipeline for BC outcome prediction using data obtained from routinely analysed blood samples, including complete blood count (CBC), white blood cell differential (DIFF), and cell population data (CPD) produced by Sysmex XN-2000 analysers.MethodsML models were trained using retrospective data produced between 2018 and 2019, from patients at Sir Charles Gairdner hospital, Nedlands, Western Australia, and processed at Pathwest Laboratory Medicine, Nedlands. Trained ML models were evaluated using stratified 10-fold cross validation.ResultsTwo ML models, an XGBoost model using CBC/DIFF/CPD features with boruta feature selection (BFS) , and a random forest model trained using CBC/DIFF features with BFS were selected for further validation after obtaining AUC scores of 0.76 pm 0.04 and 0.75 pm 0.04 respectively using stratified 10-fold cross validation. The XGBoost model obtained an AUC score of 0.76 on a internal validation set. The random forest model obtained AUC scores of 0.82 and 0.76 on internal and external validation datasets respectively.ConclusionsWe have demonstrated the utility of using an ML pipeline combined with CBC/DIFF, and CBC/DIFF/CPD feature spaces for BC outcome prediction. This builds on the growing body of research in the area of BC outcome prediction, and provides opportunity for further research.

SDPS-41 PATHOMICS AUGMENTS PREDICTIVE SURVIVAL MODELLINGIN MEDULLOBLASTOMA

Abstract PURPOSE Medulloblastoma survival and treatment response has shown wide variability based on molecular and genomic subtyping. Early prediction of overall survival (OS) may afford clinicians an opportunity to develop individualized treatment paradigms for patients. Computational pathology (“pathomics”) confers the ability to extract sub-visual histopathologic features using high-throughput analysis to better characterize tumor biology or behavior. METHODS Using the Children’s Brain Tumor Network database, pediatric patients with available clinical data and whole-slide images (WSI) were evaluated for inclusion in the study. 1mm x 1mm regions-of-interest (ROI) on each H&E-stained WSI were selected to identify tumor-rich areas for subsequent pathomic analysis, and 49 quantitative pathomics features were extracted using a digital pathology software package (Qupath). LASSO cox proportional hazard model with stratified 5-fold cross-validation was used to identify high-performing features and generate risk scores. Estimated risk scores were subsequently used to stratify patients into low-, medium- and high-risk OS cohorts. Two separate models were trained including either: 1) clinical variables only; or 2) combined clinical and pathomic features. RESULTS A total of 84 patients with median age at diagnosis of 8.61 years (range 0.31-21.72), and median OS of 45.6 months (range 0.76-195.87)) were included in the study. A survival model built using only clinical features yielded a concordance index (c-index) of 0.76 compared to actual outcomes (stratification: p<0.001). Combining clinical and high-performing pathomic features, however, yielded the best performance of the model, showing disparate outcomes between low-, medium- and high-risk groups (p<0.001) with resultant c-index of 0.85. CONCLUSION Our results highlight the utility of pathomics analysis in predicting survival in patients with pediatric medulloblastoma, which can be used to guide treatment strategy. Further work will aim to perform our analysis on an unseen validation set to improve its performance and applicability in the clinical realm.

Autism Spectrum Disorder (ASD) Identification Using Feature-Based Machine Learning Classification Model

Autism Spectrum Disorder (ASD) is a developmental disorder that impairs the development of behaviors, communication, and learning abilities. Early detection of ASD helps patients to get beter training to communicate and interact with others. In this study, we identified ASD and non-ASD individuals using machine learning (ML) approaches. We used Gaussian naive Bayes (NB), k-nearest neighbors (KNN), random forest (RF), logistic regression (LR), Gaussian naive Bayes (NB), support vector machine (SVM) with linear basis function and decision tree (DT). We preprocessed the data using the imputation methods, namely linear regression, Mice forest, and Missforest. We selected the important features using the Simultaneous perturbation feature selection and ranking (SpFSR) technique from all 21 ASD features of three datasets combined (N=1,100 individuals) from University California Irvine (UCI) repository. We evaluated the performance of the method's discrimination, calibration, and clinical utility using a stratified 10-fold cross-validation method. We achieved the highest accuracy possible by using SVM with selected the most important 10 features. We observed the integration of imputation using linear regression, SpFSR and SVM as the most effective models, with an accuracy rate of 100% outperformed the previous studies in ASD prediciton

Machine Learning Analysis of Microtensile Bond Strength of Dental Adhesives.

Dental adhesives provide retention to composite fillings in dental restorations. Microtensile bond strength (µTBS) test is the most used laboratory test to evaluate bonding performance of dental adhesives. The traditional approach for developing dental adhesives involves repetitive laboratory measurements, which consumes enormous time and resources. Machine learning (ML) is a promising tool for accelerating this process. This study aimed to develop ML models to predict the µTBS of dental adhesives using their chemical features and to identify important contributing factors for µTBS. Specifically, the chemical composition and µTBS information of 81 dental adhesives were collected from the manufacturers and the literature. The average µTBS value of each adhesive was labeled as either 0 (if <36 MPa) or 1 (if ≥36 MPa) to denote the low and high µTBS classes. The initial 9-feature data set comprised pH, HEMA, BisGMA, UDMA, MDP, PENTA, filler, fluoride, and organic solvent (OS) as input features. Nine ML algorithms, including logistic regression, k-nearest neighbor, support vector machine, decision trees and tree-based ensembles, and multilayer perceptron, were implemented for model development. Feature importance analysis identified MDP, pH, OS, and HEMA as the top 4 contributing features, which were used to construct a 4-feature data set. Grid search with stratified 10-fold cross-validation (CV) was employed for hyperparameter tunning and model performance evaluation using 2 metrics, the area under the receiver operating characteristic curve (AUC) and accuracy. The 4-feature data set generated slightly better performance than the 9-feature data set, with the highest AUC score of 0.90 and accuracy of 0.81 based on stratified CV. In conclusion, ML is an effective tool for predicting dental adhesives with low and high µTBS values and for identifying important chemical features contributing to the µTBS. The ML-based data-driven approach has great potential to accelerate the discovery of new dental adhesives and other dental materials.

Enabling new interactions with library digital collections: automatic gender recognition in historical postcards via deep learning

The Walter Havighurst Special Collections from University Archives & Preservation at Miami University's King Library has a growing collection of over 600,000 historical postcards, with approximately 30,000 digitized, primarily from the Midwest during 1890–1919. This collection supports various lines of inquiry from users, such as analyzing the evolution of gender portrayal in popular media in the United States. However, manually separating the collection into postcards of males and females would take thousands of hours, which prevents the library from supporting sociological analyses at scale. After assembling an open postcard dataset, we trained deep neural networks (i.e., YOLOv5x object detection models) to automatically detect people and classify them as male or female. Our approach limited biases in favor of one outcome by balancing the number of males and females via multi-label stratified 10-fold cross-validation. We showed that this approach can accurately detect and classify females and confidently detect and label males for the library's collection of historical postcards. Our precision of 94.9 % and recall of 33.0 % from 1890 to 1919 on male gender detection exceed the performances of 94.7 % and 31 % respectively for recognition on World War I postcards in past studies. By employing our trained deep neural networks, the library can enhance its metadata within hours and support new research inquiries at scale.

Predicting HIV infection in the decade (2005-2015) pre-COVID-19 in Zimbabwe: A supervised classification-based machine learning approach.

The burden of HIV and related diseases have been areas of great concern pre and post the emergence of COVID-19 in Zimbabwe. Machine learning models have been used to predict the risk of diseases, including HIV accurately. Therefore, this paper aimed to determine common risk factors of HIV positivity in Zimbabwe between the decade 2005 to 2015. The data were from three two staged population five-yearly surveys conducted between 2005 and 2015. The outcome variable was HIV status. The prediction model was fit by adopting 80% of the data for learning/training and 20% for testing/prediction. Resampling was done using the stratified 5-fold cross-validation procedure repeatedly. Feature selection was done using Lasso regression, and the best combination of selected features was determined using Sequential Forward Floating Selection. We compared six algorithms in both sexes based on the F1 score, which is the harmonic mean of precision and recall. The overall HIV prevalence for the combined dataset was 22.5% and 15.3% for females and males, respectively. The best-performing algorithm to identify individuals with a higher likelihood of HIV infection was XGBoost, with a high F1 score of 91.4% for males and 90.1% for females based on the combined surveys. The results from the prediction model identified six common features associated with HIV, with total number of lifetime sexual partners and cohabitation duration being the most influential variables for females and males, respectively. In addition to other risk reduction techniques, machine learning may aid in identifying those who might require Pre-exposure prophylaxis, particularly women who experience intimate partner violence. Furthermore, compared to traditional statistical approaches, machine learning uncovered patterns in predicting HIV infection with comparatively reduced uncertainty and, therefore, crucial for effective decision-making.

Predicting diabetic ketoacidosis in pediatric patients using machine learning

Background Machine learning is a powerful tool to define relationships between large data variables through computing algorithms. In medicine, machine learning can find the association between a given disease and disease-related complications such as the relationship between Diabetes and development of diabetic ketoacidosis (DKA). The aim of this study is to develop and evaluate a predicting model for diabetic ketoacidosis among pediatric cases to define the leading factors that can predict diabetic ketoacidosis. Methods We evaluated the medical records of 3737 pediatric patients between the ages of 0 and 18 years who attended diabetic clinics and were diagnosed with diabetes. After the initial data preprocessing, we used Orange, an open source software, for data visualization, and machine learning for data analysis. The study used six prediction models: Decision Tree, Random Forest, kNN, Gradient Boosting, CN2 rule inducer and AdaBoost. Data imbalance was managed using oversampling technique. Variables analyzed included age, sex, hemoglobin A1C level, visits to the diabetic education clinic, and number of appointments to diabetic clinic. Models were evaluated based on the Area under the Curve (AUC), accuracy, precision, recall and F1-score using the stratified 5-fold cross validation technique. Results The results show that the Random Forest is the highest performance classifier (AUC=0.98; F1 score=0.92; and recall=0.93). Furthermore, HbA1c was the most contributing factor to the prediction model. Conclusion This study shows the importance and effectiveness of machine learning modeling to predict the association between diabetes and the development of DKA. Flagging those patients who are at a higher risk of developing DKA provides a better point of care for these patients.

Assessment and classification of grid stability with cost-sensitive stacked ensemble classifier

Smart Grid is an intelligent power grid with a bidirectional flow of electricity and information, that applies intelligent techniques to operate the grid autonomously near the stability limit. An intelligent technique is developed to identify and predict the abnormalities due to changes in customer behaviour and the unexpected disruption in the grid. A cost-sensitive stacked ensemble classifier (CS-SEC) is proposed for predicting the operations in smart grid that combines four cost-sensitive base classifiers, namely Extreme gradient boosting, Naive Bayes, Nu-support vector machine and Random forest at level-1 and the support vector machine as the meta classifier in level-2. The meta classifier uses the probability of prediction of the first-level classifiers with stratified 5-fold cross-validation to predict the decentralized smart grid stability. The proposed stacked ensemble classifier achieved an accuracy of 98.6% with specificity, recall and precision of 98.34%, 99.0% and 99.06%, respectively. Extensive experimental evaluation and results show that the proposed CS-SEC provides an accurate prediction of grid stability compared with other state-of-the-art models. The results reveal the robustness and competency of the proposed CS-SECs with optimized parameters.

Deep learning-based histomorphological pattern profiles for effective risk stratification in prostate cancer.

e17107 Background: Prostate cancer affects millions of men globally and biochemical recurrence (BCR) is an important indicator of its progression. With the increasing use of computational techniques, there has been a growing interest in utilizing whole slide images (WSIs) of H&amp;E stained prostate tissue to evaluate the patients’ clinical risks. Our study aimed to develop a novel method for predicting BCR by performing clustering on features extracted from deep learning models, thereby capturing morphological characteristics. Methods: The WSIs and clinical information were collected from The Cancer Genome Atlas (TCGA, from 2000 to 2013) dataset, for 321 patients (43 patients with BCR) who were not treated with any adjuvant therapies, with 2.5 years of median follow-up time. 356 collected WSIs were split into 1024 x 1024 patches at 5x magnification for the deep learning-based analysis. We used an ImageNet-pretrained deep learning-based clustering model (SwAV) to extract features from each patch and applied additional K-means clustering to assign the patches into one of 16 clusters. We tested both the cosine similarity and Euclidean distance as the clustering measure. After excluding the dissimilar patches under specific thresholds, the proportion of each cluster was calculated per slide to describe the distribution of morphological patterns as a 16-dimensional vector. To evaluate the predictability, we trained Cox models with the proportion vectors and assessed their performance via the concordance index (c-index). We repeated a stratified 5-fold cross-validation process 100 times to achieve robustness. As a reference, we conducted the same experiment with a BCR prediction nomogram published by MSKCC. Moreover, with the median proportion of each cluster, we separated two patient groups, drawed Kaplan-Meier curves, performed the log-rank test for significance, and we obtained the hazard ratio (HR) of each cluster from the trained Cox model of each fold. Results: We explored several thresholds to find ones yielding the best performance. When using the similarity, the best c-index was 0.704 with the threshold value of 0.25. With the distance value of 0.60, the c-index was 0.701. When using the MSKCC nomogram that incorporates clinical features, the average c-index was 0.717. The table contains the results of statistical analysis for selected clusterings that were significant in the log-rank test. Particularly in cluster B, the patches showed high-risk patterns including perineural invasion, high grade cancer, and necrosis in agreement with the high HR. Conclusions: Our results suggest that the histomorphological pattern profiles depicted by the deep learning model have the potential for effective risk stratification in prostate cancer. [Table: see text]

Development, Evaluation, and Application of Machine Learning Models for Accurate Prediction of Root Uptake of Per- and Polyfluoroalkyl Substances.

Machine learning (ML) models were developed for understanding the root uptake of per- and polyfluoroalkyl substances (PFASs) under complex PFAS-crop-soil interactions. Three hundred root concentration factor (RCF) data points and 26 features associated with PFAS structures, crop properties, soil properties, and cultivation conditions were used for the model development. The optimal ML model, obtained by stratified sampling, Bayesian optimization, and 5-fold cross-validation, was explained by permutation feature importance, individual conditional expectation plot, and 3D interaction plot. The results showed that soil organic carbon contents, pH, chemical logP, soil PFAS concentration, root protein contents, and exposure time greatly affected the root uptake of PFASs with 0.43, 0.25, 0.10, 0.05, 0.05, and 0.05 of relative importance, respectively. Furthermore, these factors presented the key threshold ranges in favor of the PFAS uptake. Carbon-chain length was identified as the critical molecular structure affecting root uptake of PFASs with 0.12 of relative importance, based on the extended connectivity fingerprints. A user-friendly model was established with symbolic regression for accurately predicting RCF values of the PFASs (including branched PFAS isomerides). The present study provides a novel approach for profound insight into the uptake of PFASs by crops under complex PFAS-crop-soil interactions, aiming to ensure food safety and human health.

Evaluation of CNN algorithm at locomotion mode identification for a knee-assisted exoskeleton

In this study, a knee-assisted exoskeleton that can assist the wearer's knee joints at the appropriate phase of the gait cycle is designed. A convolutional neural network (CNN) model is used in the gait control to identify the locomotion modes of the exoskeleton. In order to verify the CNN-based locomotion mode identification method, an offline CNN model was created in MATLAB. The experiment was conducted and seven healthy subjects’ hip, knee, and ankle joint angles were captured and six common locomotion modes were recognized in the experiments. The recorded data were segmented with overlapping sliding windows of varying sizes and were tested using a stratified 5-fold cross-validation technique. The experimental findings showed that the appropriate window size was 1.25T (T stands for the normal adult gait cycle). The lowest and the highest accuracies of the test sets at the window size of 1.25T were 96.47% ± 0.83% and 99.67% ± 0.49% on different subjects. It also showed that the model has strong generalization. Thus, the algorithm had proved useful to extract the characteristics of lower limb joint angles in various locomotion modes and the approach in this paper can be extended to the recognition of motion patterns in lower limb exoskeletons.

Analysis of Accuracy Metric of Machine Learning Algorithms in Predicting Heart Disease

Introduction: Heart disease is, for the most part, alluding to conditions that include limited or blocked veins that can prompt a heart attack, chest torment or stroke. Earlier identification of heart disease may reduce the death rate. The cost of medical diagnosis makes it perverse to cure it for the large amount of people early. Using machine learning models performed on dataset. This article aims to find the most efficient and accurate machine learning models for disease prediction.Material and Methods: Several supervised machine learning algorithms were utilized to diagnosis and prediction of heart disease such as logistic regression, decision tree, random forest and KNN. The algorithms are applied to a dataset taken from the Kaggle site including 70000 samples. In algorithms, methods such as the importance of features, hold out validation, 10-fold cross-validation, stratified 10-fold cross-validation, leave one out cross-validation are the result of effective performance and increase accuracy. In addition, feature importance scores was estimated for each feature in some algorithms. These features were ranked based on feature importance score. All the work is done in the Anaconda environment based on python programming language and Scikit-learn library.Results: The algorithms performance is compared to each other so that performance based on ROC curve and some criteria such as accuracy, precision, sensitivity and F1 score were evaluated for each model. As a result of evaluation, random forest algorithm with F1 score 92%, accuracy 92% and AUC ROC 95%, has better performance than other algorithms.Conclusion: The area under the ROC curve and evaluating criteria related to a number of classifying algorithms of machine learning to evaluate heart disease and indeed, the diagnosis and prediction of heart disease is compared to determine the most appropriate classifier.

Real-Time Prediction of Sepsis in Critical Trauma Patients: Machine Learning–Based Modeling Study

Sepsis is a leading cause of death in patients with trauma, and the risk of mortality increases significantly for each hour of delay in treatment. A hypermetabolic baseline and explosive inflammatory immune response mask clinical signs and symptoms of sepsis in trauma patients, making early diagnosis of sepsis more challenging. Machine learning-based predictive modeling has shown great promise in evaluating and predicting sepsis risk in the general intensive care unit (ICU) setting, but there has been no sepsis prediction model specifically developed for trauma patients so far. To develop a machine learning model to predict the risk of sepsis at an hourly scale among ICU-admitted trauma patients. We extracted data from adult trauma patients admitted to the ICU at Beth Israel Deaconess Medical Center between 2008 and 2019. A total of 42 raw variables were collected, including demographics, vital signs, arterial blood gas, and laboratory tests. We further derived a total of 485 features, including measurement pattern features, scoring features, and time-series variables, from the raw variables by feature engineering. The data set was randomly split into 70% for model development with stratified 5-fold cross-validation, 15% for calibration, and 15% for testing. An Extreme Gradient Boosting (XGBoost) model was developed to predict the hourly risk of sepsis at prediction windows of 4, 6, 8, 12, and 24 hours. We evaluated model performance for discrimination and calibration both at time-step and outcome levels. Clinical applicability of the model was evaluated with varying levels of precision, and the potential clinical net benefit was assessed with decision curve analysis (DCA). A Shapley additive explanation algorithm was applied to show the effect of features on the prediction model. In addition, we trained an L2-regularized logistic regression model to compare its performance with XGBoost. We included 4603 trauma patients in the study, 1196 (26%) of whom developed sepsis. The XGBoost model achieved an area under the receiver operating characteristics curve (AUROC) ranging from 0.83 to 0.88 at the 4-to-24-hour prediction window in the test set. With a ratio of 9 false alerts for every true alert, it predicted 73% (386/529) of sepsis-positive timesteps and 91% (163/179) of sepsis events in the subsequent 6 hours. The DCA showed our model had a positive net benefit in the threshold probability range of 0 to 0.6. In comparison, the logistic regression model achieved lower performance, with AUROC ranging from 0.76 to 0.84 at the 4-to-24-hour prediction window. The machine learning-based model had good discrimination and calibration performance for sepsis prediction in critical trauma patients. Using the model in clinical practice might help to identify patients at risk of sepsis in a time window that enables personalized intervention and early treatment.

Research on prediction of e-commerce repurchase behavior based on multiple fusion models

With the advent of the Internet era, online shopping has become an integral part of peoples life. In order to perform precision marketing, more and more e-commerce platforms are trying to predict users repurchase behaviors by collecting massive user behavior data. Although the traditional single-model prediction method is mature, it is still difficult to improve the accuracy of prediction. Based on the real user behavior data of Tmall, this paper focuses on comparing and exploring the help of different algorithm fusion methods to improve the model prediction effect. The under-sampling method is introduced for sample equalization processing. User behavior features are constructed from three aspects which are user, merchant and user-merchant interaction. Taking AUC value as evaluation method, Soft-Voting and Stacking model fusion methods are used to integrate logistics regression, KNN, XGBoost and RandomForest. And the prediction results is produced based on stratified 5-fold cross-validation. The experimental results show that the fusion model can effectively improve the prediction effect, and the AUC value is raised by 0.2%~4% compared with the single model. The AUC value of Soft-Voting increases by approximately 0.4% after it is weighted.

Automatic signal quality assessment of raw trans-abdominal biopotential recordings for non-invasive fetal electrocardiography.

Introduction: Wearable monitoring systems for non-invasive multi-channel fetal electrocardiography (fECG) can support fetal surveillance and diagnosis during pregnancy, thus enabling prompt treatment. In these embedded systems, power saving is the key to long-term monitoring. In this regard, the computational burden of signal processing methods implemented for the fECG extraction from the multi-channel trans-abdominal recordings plays a non-negligible role. In this work, a supervised machine-learning approach for the automatic selection of the most informative raw abdominal recordings in terms of fECG content, i.e., those potentially leading to good-quality, non-invasive fECG signals from a low number of channels, is presented and evaluated. Methods: For this purpose, several signal quality indexes from the scientific literature were adopted as features to train an ensemble tree classifier, which was asked to perform a binary classification between informative and non-informative abdominal channels. To reduce the dimensionality of the classification problem, and to improve the performance, a feature selection approach was also implemented for the identification of a subset of optimal features. 10336 5-s long signal segments derived from a real dataset of multi-channel trans-abdominal recordings acquired from 55 voluntary pregnant women between the 21st and the 27th week of gestation, with healthy fetuses, were adopted to train and test the classification approach in a stratified 10-time 10-fold cross-validation scheme. Abdominal recordings were firstly pre-processed and then labeled as informative or non-informative, according to the signal-to-noise ratio exhibited by the extracted fECG, thus producing a balanced dataset of bad and good quality abdominal channels. Results and Discussion: Classification performance revealed an accuracy above 86%, and more than 88% of those channels labeled as informative were correctly identified. Furthermore, by applying the proposed method to 50 annotated 24-channel recordings from the NInFEA dataset, a significant improvement was observed in fetal QRS detection when only the channels selected by the proposed approach were considered, compared with the use of all the available channels. As such, our findings support the hypothesis that performing a channel selection by looking directly at the raw abdominal signals, regardless of the fetal presentation, can produce a reliable measurement of fetal heart rate with a lower computational burden.

Prediction of Suicidal Behaviors in the Middle-aged Population: Machine Learning Analyses of UK Biobank.

Suicidal behaviors, including suicide deaths and attempts, are major public health concerns. However, previous suicide models required a huge amount of input features, resulting in limited applicability in clinical practice. We aimed to construct applicable models (ie, with limited features) for short- and long-term suicidal behavior prediction. We further validated these models among individuals with different genetic risks of suicide. Based on the prospective cohort of UK Biobank, we included 223 (0.06%) eligible cases of suicide attempts or deaths, according to hospital inpatient or death register data within 1 year from baseline and randomly selected 4460 (1.18%) controls (1:20) without such records. We similarly identified 833 (0.22%) cases of suicidal behaviors 1 to 6 years from baseline and 16,660 (4.42%) corresponding controls. Based on 143 input features, mainly including sociodemographic, environmental, and psychosocial factors; medical history; and polygenic risk scores (PRS) for suicidality, we applied a bagged balanced light gradient-boosting machine (LightGBM) with stratified 10-fold cross-validation and grid-search to construct the full prediction models for suicide attempts or deaths within 1 year or between 1 and 6 years. The Shapley Additive Explanations (SHAP) approach was used to quantify the importance of input features, and the top 20 features with the highest SHAP values were selected to train the applicable models. The external validity of the established models was assessed among 50,310 individuals who participated in UK Biobank repeated assessments both overall and by the level of PRS for suicidality. Individuals with suicidal behaviors were on average 56 years old, with equal sex distribution. The application of these full models in the external validation data set demonstrated good model performance, with the area under the receiver operating characteristic (AUROC) curves of 0.919 and 0.892 within 1 year and between 1 and 6 years, respectively. Importantly, the applicable models with the top 20 most important features showed comparable external-validated performance (AUROC curves of 0.901 and 0.885) as the full models, based on which we found that individuals in the top quintile of predicted risk accounted for 91.7% (n=11) and 80.7% (n=25) of all suicidality cases within 1 year and during 1 to 6 years, respectively. We further obtained comparable prediction accuracy when applying these models to subpopulations with different genetic susceptibilities to suicidality. For example, for the 1-year risk prediction, the AUROC curves were 0.907 and 0.885 for the high (>2nd tertile of PRS) and low (<1st) genetic susceptibilities groups, respectively. We established applicable machine learning-based models for predicting both the short- and long-term risk of suicidality with high accuracy across populations of varying genetic risk for suicide, highlighting a cost-effective method of identifying individuals with a high risk of suicidality.

Intravenous immunoglobulin resistance in Kawasaki disease patients: prediction using clinical data

BackgroundAbout 10–20% of Kawasaki disease (KD) patients are resistant to the initial infusion of intravenous immunoglobin (IVIG). The aim of this study was to assess whether IVIG resistance in KD patients could be predicted using standard clinical and laboratory features.MethodsData were from two cohorts: a Korean cohort of 7101 KD patients from 2015 to 2017 and a cohort of 649 KD patients from San Diego enrolled from 1998 to 2021. Features included laboratory values, the worst Z-score from the initial echocardiogram or during hospitalization, and the five clinical KD signs at presentation.ResultsFive machine learning models achieved a maximum median AUC of 0.711 [IQR: 0.706–0.72] in the Korean cohort and 0.696 [IQR: 0.609–0.722] in the San Diego cohort during stratified 10-fold cross-validation using significant laboratory features identified from univariate analysis. Adding the Z-score, KD clinical signs, or both did not considerably improve the median AUC in either cohort.ConclusionsUsing commonly measured clinical laboratory data alone or in conjunction with echocardiographic findings and clinical features is not sufficient to predict IVIG resistance. Further attempts to predict IVIG resistance will need to incorporate additional data such as transcriptomics, proteomics, and genetics to achieve meaningful predictive utility.ImpactWe demonstrated that laboratory, echocardiographic, and clinical findings cannot predict intravenous immunoglobin (IVIG) resistance to a clinically meaningful extent using machine learning in a homogenous Asian or ethnically diverse population of patients with Kawasaki disease (KD).Visualizing these features using uniform manifold approximation and projection (UMAP) is an important step to evaluate predictive utility in a qualitative manner.Further attempts to predict IVIG resistance in KD patients will need to incorporate novel biomarkers or other specialized features such as genetic differences or transcriptomics to be clinically useful.