Ensemble Voting Research Articles

A predictive machine-learning model for clinical decision-making in washed microbiota transplantation on ulcerative colitis

Background and AimMachine learning based on clinical data and treatment protocols for better clinical decision-making is a current research hotspot. This study aimed to build a machine learning model on washed microbiota transplantation (WMT) for ulcerative colitis (UC), providing patients and clinicians with a new evaluation system to optimize clinical decision-making.MethodsPatients with UC who underwent WMT via mid-gut or colonic delivery route at an affiliated hospital of Nanjing Medical University from April 2013 to June 2022 were recruited. Model ensembles based on the clinical indicators were constructed by machine-learning to predict the clinical response of WMT after one month.ResultsA total of 366 patients were enrolled in this study, with 210 patients allocated for training and internal validation, and 156 patients for external validation. The low level of indirect bilirubin, activated antithrombin III, defecation frequency and cholinesterase and the elderly and high level of creatine kinase, HCO3- and thrombin time were related to the clinical response of WMT at one month. Besides, the voting ensembles exhibited an area under curve (AUC) of 0.769 ± 0.019 [accuracy, 0.754; F1-score, 0.845] in the internal validation; the AUC of the external validation was 0.614 ± 0.017 [accuracy, 0.801; F1-score, 0.887]. Additionally, the model was available at https://wmtpredict.streamlit.app.ConclusionsThis study pioneered the development of a machine learning model to predict the one-month clinical response of WMT on UC. The findings demonstrate the potential value of machine learning applications in the field of WMT, opening new avenues for personalized treatment strategies in gastrointestinal disorders.Trial registrationclinical trials, NCT01790061. Registered 09 February 2013 - Retrospectively registered, https://clinicaltrials.gov/study/NCT01790061

Improving Cleaning of Solar Systems through Machine Learning Algorithms

The study focuses on the importance of maintaining photovoltaic (PV) systems for optimal performance in sustainable energy generation. It highlights the impact of dust accumulation on reducing system efficiency and proposes a method to predict system performance, aiding in scheduling cleaning activities effectively. Two prediction models are developed: one using time-series prediction techniques (LSTM, ARIMA, SARIMAX) to forecast Performance Ratio (PR), and another employing ensemble voting classifiers (RF, Log, GBM) to predict the need for cleaning. The SARIMAX model performs best, achieving high accuracy in PR prediction (R2 = 92.12%), while the classification model accurately predicts cleaning needs (91%). The research provides valuable insights for improving maintenance strategies and enhancing the efficiency and sustainability of PV systems.

Machine-learning-based ensemble regression for vehicle-to-vehicle distance estimation using a toe-in style stereo camera

Adjusting the following distance from the front vehicle in highway traffic is important to reduce the risk of collision. Distance estimation is an important research area for advanced driver assistance systems. Therefore, this paper presents a methodology that combines the strengths of several machine learning algorithms using joint decision mechanisms and searches for optimal results for vehicle-to-vehicle distance estimation. The hyperparameter optimization of machine learning models is performed by an iterative algorithm that compares combinations of hyperparameter values. In addition, machine learning algorithms are combined and tested with ensemble learning methods to improve the results obtained. According to the experiments, the ensemble voting regression created by combining extreme gradient boosting, categorical boosting and two multi-layer perceptron models achieves the best result with a mean absolute percentage error value of 1.5444. Considering the comparisons made with other methods, the accuracy of the results obtained with the proposed method is quite high.Code is at: https://github.com/ozgurduran/V2V-VR.

Achieving High Accuracy in Android Malware Detection through Genetic Programming Symbolic Classifier

The detection of Android malware is of paramount importance for safeguarding users’ personal and financial data from theft and misuse. It plays a critical role in ensuring the security and privacy of sensitive information on mobile devices, thereby preventing unauthorized access and potential damage. Moreover, effective malware detection is essential for maintaining device performance and reliability by mitigating the risks posed by malicious software. This paper introduces a novel approach to Android malware detection, leveraging a publicly available dataset in conjunction with a Genetic Programming Symbolic Classifier (GPSC). The primary objective is to generate symbolic expressions (SEs) that can accurately identify malware with high precision. To address the challenge of imbalanced class distribution within the dataset, various oversampling techniques are employed. Optimal hyperparameter configurations for GPSC are determined through a random hyperparameter values search (RHVS) method developed in this research. The GPSC model is trained using a 10-fold cross-validation (10FCV) technique, producing a set of 10 SEs for each dataset variation. Subsequently, the most effective SEs are integrated into a threshold-based voting ensemble (TBVE) system, which is then evaluated on the original dataset. The proposed methodology achieves a maximum accuracy of 0.956, thereby demonstrating its effectiveness for Android malware detection.

Transfer Learning Artificial Neural Network-based Ensemble Voting of Water Quality Classification for Different Types of Farming

Transfer Learning Artificial Neural Network-based Ensemble Voting of Water Quality Classification for Different Types of Farming

Enhance the Detection of DoS and Brute Force Attacks within the MQTT Environment Through Feature Engineering and Employing an Ensemble Technique

The rapid development of the Internet of Things (IoT) environment has introduced unprecedented levels of connectivity and automation. The Message Queuing Telemetry Transport (MQTT) protocol has become recognized in IoT applications due to its lightweight and efficient features; however, this simplicity also renders MQTT vulnerable to multiple attacks that can be launched against the protocol, including denial of service (DoS) and brute-force attacks. This study aims to improve the detection of intrusion DoS and bruteforce attacks in an MQTT traffic intrusion detection system (IDS). Our approach utilizes the MQTT dataset for model training by employing effective feature engineering and ensemble learning techniques. Following our analysis and comparison, we identified the top 10 features demonstrating the highest effectiveness, leading to improved model accuracy. We used supervised machine learning models, including Random Forest, Decision Trees, k-Nearest Neighbors, and XGBoost, in combination with ensemble classifiers. Stacking, voting, and bagging ensembles utilize these four supervised machinelearning methods to combine models. This study's results illustrate the proposed technique's efficacy in enhancing the accuracy of detecting DoS and brute-force attacks in MQTT traffic. Stacking and voting classifiers achieved the highest accuracy of 0.9538. Our approach outperforms the most recent study that utilized the same dataset.

Utilising deep learning networks to classify ZEB2 expression images in cervical cancer.

Aims/Background Cervical cancer continues to be a significant cause of cancer-related deaths among women, especially in low-resource settings where screening and follow-up care are lacking. The transcription factor zinc finger E-box-binding homeobox 2 (ZEB2) has been identified as a potential marker for tumour aggressiveness and cancer progression in cervical cancer tissues. Methods This study presents a hybrid deep learning system developed to classify cervical cancer images based on ZEB2 expression. The system integrates multiple convolutional neural network models-EfficientNet, DenseNet, and InceptionNet-using ensemble voting. We utilised the gradient-weighted class activation mapping (Grad-CAM) visualisation technique to improve the interpretability of the decisions made by the convolutional neural networks. The dataset consisted of 649 annotated images, which were divided into training, validation, and testing sets. Results The hybrid model exhibited a high classification accuracy of 94.4% on the test set. The Grad-CAM visualisations offered insights into the model's decision-making process, emphasising the image regions crucial for classifying ZEB2 expression levels. Conclusion The proposed hybrid deep learning model presents an effective and interpretable method for the classification of cervical cancer based on ZEB2 expression. This approach holds the potential to substantially aid in early diagnosis, thereby potentially enhancing patient outcomes and mitigating healthcare costs. Future endeavours will concentrate on enhancing the model's accuracy and investigating its applicability to other cancer types.

Intrusion Detection System with Ensemble Machine Learning Approaches using VotingClassifier

Internets have become a part of our everyday life due to the advancement in the electronics and signal processing technologies during past decades. The tremendous growth of internet leads towards the network threats. Many times firewalls and anti-viruses fails to manage the network because of this Intrusion Detection System (IDS) comes to assists us. In this paper we use IDS with Ensemble methodologies utilized in machine learning involve the fusion of multiple classifiers to improve predictive performance, while voting classifiers combine predictions from individual models to reach conclusive decisions. The paper employs a voting ensemble method combing decision tree, logistic regression and support vector machine classifier models. We test our proposedmodel to classify the NSL-KDD dataset. Our ensemble methodologies of proposed algorithmproduce a good result.

Development of an efficient novel method for coronary artery disease prediction using machine learning and deep learning techniques.

Heart disease is a severe health issue that results in high fatality rates worldwide. Identifying cardiovascular diseases such as coronary artery disease (CAD) and heart attacks through repetitive clinical data analysis is a significant task. Detecting heart disease in its early stages can save lives. The most lethal cardiovascular condition is CAD, which develops over time due to plaque buildup in coronary arteries, causing incomplete blood flow obstruction. Machine Learning (ML) is progressively used in the medical sector to detect CAD disease. The primary aim of this work is to deliver a state-of-the-art approach to enhancing CAD prediction accuracy by using a DL algorithm in a classification context. A unique ML technique is proposed in this study to predict CAD disease accurately using a deep learning algorithm in a classification context. An ensemble voting classifier classification model is developed based on various methods such as Naïve Bayes (NB), Logistic Regression (LR), Decision Tree (DT), XGBoost, Random Forest (RF), Convolutional Neural Network (CNN), Support Vector Machine (SVM), K Nearest Neighbor (KNN), Bidirectional LSTM and Long Short-Term Memory (LSTM). The performance of the ensemble models and a novel model are compared in this study. The Alizadeh Sani dataset, which consists of a random sample of 216 cases with CAD, is used in this study. Synthetic Minority Over Sampling Technique (SMOTE) is used to address the issue of imbalanced datasets, and the Chi-square test is used for feature selection optimization. Performance is assessed using various assessment methodologies, such as confusion matrix, accuracy, recall, precision, f1-score, and auc-roc. When a novel algorithm achieves the highest accuracy relative to other algorithms, it demonstrates its effectiveness in several ways, including superior performance, robustness, generalization capability, efficiency, innovative approaches, and benchmarking against baselines. These characteristics collectively contribute to establishing the novel algorithm as a promising solution for addressing the target problem in machine learning and related fields. Implementing the novel model in this study significantly improved performance, achieving a prediction accuracy rate of 92% in the detection of CAD. These findings are competitive and on par with the top outcomes among other methods.

Temporal-Quality Ensemble Technique for Handling Image Blur in Packaging Defect Inspection.

Despite achieving numerous successes with surface defect inspection based on deep learning, the industry still faces challenges in conducting packaging defect inspections that include critical information such as ingredient lists. In particular, while previous achievements primarily focus on defect inspection in high-quality images, they do not consider defect inspection in low-quality images such as those containing image blur. To address this issue, we proposed a noble inference technique named temporal-quality ensemble (TQE), which combines temporal and quality weights. Temporal weighting assigns weights to input images by considering the timing in relation to the observed image. Quality weight prioritizes high-quality images to ensure the inference process emphasizes clear and reliable input images. These two weights improve both the accuracy and reliability of the inference process of low-quality images. In addition, to experimentally evaluate the general applicability of TQE, we adopt widely used convolutional neural networks (CNNs) such as ResNet-34, EfficientNet, ECAEfficientNet, GoogLeNet, and ShuffleNetV2 as the backbone network. In conclusion, considering cases where at least one low-quality image is included, TQE has an F1-score approximately 17.64% to 22.41% higher than using single CNN models and about 1.86% to 2.06% higher than an average voting ensemble.

The object detection model uses combined extraction with KNN and RF classification

Object detection plays an important role in various fields. Developing detection models for 2D objects that experience rotation and texture variations is a challenge. In this research, the initial stage of the proposed model integrates the gray-level co-occurrence matrix (GLCM) and local binary patterns (LBP) texture feature extraction to obtain feature vectors. The next stage is classifying features using k-nearest neighbors (KNN) and random forest (RF), as well as voting ensemble (VE). System testing used a dataset of 4,437 2D images, the results for KNN accuracy were 92.7% and F1-score 92.5%, while RF performance was lower. Although GLCM features improve performance on both algorithms, KNN is more consistent. The VE approach provides the best performance with an accuracy of 93.9% and an F1-score of 93.8%, this shows the effectiveness of the ensemble technique in increasing object detection accuracy. This study contributes to the field of object detection with a new approach combining GLCM and LBP as feature vectors as well as VE for classification.

Identifying First-Trimester Risk Factors for SGA-LGA Using Weighted Inheritance Voting Ensemble Learning.

The classification of fetuses as Small for Gestational Age (SGA) and Large for Gestational Age (LGA) is a critical aspect of neonatal health assessment. SGA and LGA, terms used to describe fetal weights that fall below or above the expected weights for Appropriate for Gestational Age (AGA) fetuses, indicate intrauterine growth restriction and excessive fetal growth, respectively. Early prediction and assessment of latent risk factors associated with these classifications can facilitate timely medical interventions, thereby optimizing the health outcomes for both the infant and the mother. This study aims to leverage first-trimester data to achieve these objectives. This study analyzed data from 7943 pregnant women, including 424 SGA, 928 LGA, and 6591 AGA cases, collected from 2015 to 2021 at the Third Affiliated Hospital of Sun Yat-sen University in Guangzhou, China. We propose a novel algorithm, named the Weighted Inheritance Voting Ensemble Learning Algorithm (WIVELA), to predict the classification of fetuses into SGA, LGA, and AGA categories based on biochemical parameters, maternal factors, and morbidity during pregnancy. Additionally, we proposed algorithms for relevance determination based on the classifier to ascertain the importance of features associated with SGA and LGA. The proposed classification solution demonstrated a notable average accuracy rate of 92.12% on 10-fold cross-validation over 100 loops, outperforming five state-of-the-art machine learning algorithms. Furthermore, we identified significant latent maternal risk factors directly associated with SGA and LGA conditions, such as weight change during the first trimester, prepregnancy weight, height, age, and obstetric factors like fetal growth restriction and birthing LGA baby. This study also underscored the importance of biomarker features at the end of the first trimester, including HDL, TG, OGTT-1h, OGTT-0h, OGTT-2h, TC, FPG, and LDL, which reflect the status of SGA or LGA fetuses. This study presents innovative solutions for classifying and identifying relevant attributes, offering valuable tools for medical teams in the clinical monitoring of fetuses predisposed to SGA and LGA conditions during the initial stage of pregnancy. These proposed solutions facilitate early intervention in nutritional care and prenatal healthcare, thereby contributing to enhanced strategies for managing the health and well-being of both the fetus and the expectant mother.

Rainfall classification and forecasting based on a novel voting adaptive dynamic optimization algorithm

Environmental issues of rainfall are basic in terms of understanding and management of ecosystems and natural resources. The rainfall patterns significantly affect soil moisture, vegetation growth and biodiversity in the ecosystems. In addition, proper classification of rainfall types helps in the evaluation of the risk of flood, drought, and other extreme weather events’ risk, which immensely affect the ecosystems and human societies. Rainfall classification can be improved by using machine learning and metaheuristic algorithms. In this work, an Adaptive Dynamic Puma Optimizer (AD-PO) algorithm combined with Guided Whale Optimization Algorithm (Guided WOA) introduces a potentially important improvement in rainfall classification approaches. These algorithms are to be combined to enable researchers to comprehend and classify rain events by their specific features, such as intensity, duration, and spatial distribution. A voting ensemble approach within the proposed (AD-PO-Guided WOA) algorithm increases its predictive performance because of the combination of predictions from several classifiers to localize the dominant rainfall class. The presented approach not only makes the classifying of rain faster and more accurate but also strengthens the robustness and trustworthiness of the classification in this regard. Comparison to other optimization algorithms validates the effectiveness of the AD-PO-Guided WOA algorithm in terms of performance metrics with an outstanding 95.99% accuracy. Furthermore, the second scenario is applied for forecasting based on the long short-term memory networks (LSTM) model optimized by the AD-PO-Guided WOA algorithm. The AD-PO-Guided WOA- LSTM algorithm produces rainfall prediction with an MSE of 0.005078. Wilcoxon rank test, descriptive statistics, and sensitivity analysis are applied to help evaluating and improving the quality and validity of the proposed algorithm. This intensive method facilitates rainfall classification and is a base for suggested measures that cut the hazards of extreme weather events on societies.

Crop Recommendation System Using Machine Learning: A Comparative Study

Agriculture, as a fundamental aspect of human existence, faces challenges in crop selection, impacting resource allocation and productivity. This project addresses these challenges by proposing a stable system employing a soft voting classifier ensemble method. The ensemble comprises Naive Bayes, Support Vector Machine (SVM), Decision Tree, and Random Forest classifiers, offering personalized crop recommendations. Feasibility analysis encompasses technical, operational, economic, and scheduling aspects, ensuring practicality and efficacy. Development follows an incremental model, emphasizing continuous enhancement through feedback. Results indicate accuracies for individual classifiers (’Decision Tree’: 98.38%, ’Random Forest’: 98.90%, ’Naive Bayes’: 98.14%, ’SVM’: 98.50%), with an ensemble accuracy of 98.99%. Cross-validation confirms robustness. Evaluation metrics such as recall, precision, and F1 score demonstrate that the soft voting ensemble outperforms individual classifiers, highlighting its effectiveness in optimizing crop selection processes in agriculture and facilitating improved resource management and productivity.

A Novel Dataset and Approach for Adversarial Attack Detection in Connected and Automated Vehicles

Adversarial attacks have received much attention as communication network applications rise in popularity. Connected and Automated Vehicles (CAVs) must be protected against adversarial attacks to ensure passenger and vehicle safety on the road. Nevertheless, CAVs are susceptible to several types of attacks, such as those that target intra- and inter-vehicle networks. These harmful attacks not only cause user privacy and confidentiality to be lost, but they also have more grave repercussions, such as physical harm and death. It is critical to precisely and quickly identify adversarial attacks to protect CAVs. This research proposes (1) a new dataset comprising three adversarial attacks in the CAV network traffic and normal traffic, (2) a two-phased adversarial attack detection technique named TAAD-CAV, where in the first phase, an ensemble voting classifier having three machine learning classifiers and one separate deep learning classifier is trained, and the output is used in the next phase. In the second phase, a meta classifier (i.e., Decision Tree is used as a meta classifier) is trained on the combined predictions from the previous phase to detect adversarial attacks. We preprocess the dataset by cleaning data, removing missing values, and adjusting the Z-score normalization. Evaluation metrics such as accuracy, recall, precision, F1-score, and confusion matrix are employed to evaluate and compare the performance of the proposed model. Results reveal that TAAD-CAV achieves the highest accuracy with a value of 70% compared with individual ML and DL classifiers.

Classification of Diabetes Using Ensemble Machine Learning Techniques

Diabetes is a widespread chronic condition that impacts people all over the globe and requires a clear and timely diagnosis. Untreated diabetes leads to retinopathy, nephropathy, and damage to the nervous system. In this context, Machine Learning (ML) might be used to detect health problems early, diagnose them, and track their progress. Ensemble techniques are a promising approach that combines many classifiers to improve forecast accuracy and resilience. This study investigates the categorization of diabetes using an ensemble machine learning technique known as a voting classifier. Using a variety of classifiers, including Light Gradient Boosting Machine (LightGBM), Gradient Boost classifier (GBC), and Random Forest (RF). The predictions are aggregated using voting methods to get a final classification result. The research is carried out using two benchmarking datasets: the Pima Indian Diabetes Dataset (PIDD) and the German Dataset. The Boruta technique is used to choose the best attributes from the datasets, while the Random Over Sampling approach balances the range of classes and eliminates abnormal data using the interquartile range approach. The findings showed that the combination of the Boruta feature selection algorithm and ensemble Voting Classifier performed better for both PIDD and German datasets with an accuracy of 93% and 90% respectively. These algorithms are evaluated and the maximum accuracy is produced using the combination of the Boruta feature selection algorithm and ensemble Voting Classifier. This research helps medical professionals in the early prediction of diabetes, reducing physician’s time.

Regressor cascading for time series forecasting

Time series forecasting is the process of predicting future values of a time series based on its historical data patterns. It is a critical task in many domains, including finance, supply chain management, the environment, and more as accurate forecasts can help businesses and organizations make better decisions and improve their metrics. Although there have been significant advances in time series forecasting systems, thanks to the development of new machine learning algorithms, hardware improvements, and the increasing availability of data, it remains a challenging task. Common pitfalls, especially of single-model approaches include susceptibility to noise and outliers and inability to handle non-stationary data, which can lead to inaccurate and non-robust forecasts. Model-combining approaches, such as averaging the results of multiple predictors to produce a final forecast, are commonly used to mitigate such issues. This work introduces a novel application of Cascade Generalization or Cascading for time series forecasting, where multiple predictors are used sequentially, with each predictor’s output serving as additional input for the next. This methodology aims to overcome the limitations of single-model forecasts and traditional ensembles by incorporating a progressive learning mechanism. We adapt Cascade Generalization specifically for time series data, detailing its implementation and potential for handling complex, dynamic datasets. Our approach was systematically evaluated against traditional two-model averaging ensembles across ten diverse datasets, employing the Root Mean Square Error (RMSE) metric for performance assessment. The results revealed that cascading tends to outperform voting ensembles in most cases. This consistent trend suggests that cascading can be considered a reliable alternative to voting ensembles, showcasing its potential as an effective strategy for improving time series forecasting across a wide range of scenarios.

An efficient synthetic minority oversampling technique-based ensemble learning model to detect COVID-19 severity

The COVID-19 pandemic has highlighted the importance of accurately predicting disease severity to ensure timely intervention and effective allocation of healthcare resources, which can ultimately improve patient outcomes. This study aims to develop an efficient machine learning (ML) model based on patient demographic and clinical data. It utilizes advanced feature engineering techniques to reduce the dimensionality of dataset and address the issue of highly imbalanced data using synthetic minority oversampling technique (SMOTE). The study employs several ensemble learning models, including XGBoost, Random Forest, AdaBoost, voting ensemble, enhanced-weighted voting ensemble, and stack-based ensembles with support vector machine (SVM) and Gaussian Naïve Bayes as meta-learners, to develop the proposed model. The results indicate that the proposed model outperformed the top-performing models reported in previous studies. It achieved an accuracy of 0.978, sensitivity of 1.0, precision of 0.875, F1-score of 0.934, and receiver operating characteristic area under the curve (ROC-AUC) of 0.965. The study identified several features that significantly correlated with COVID-19 severity, which included respiratory rate (breaths per minute), c-reactive proteins, age, and total leukocyte count (TLC) count. The proposed approach presents a promising method for accurate COVID-19 severity prediction, which may prove valuable in assisting healthcare providers in making informed decisions about patient care.

Combined Use of Multiple Cloud‐Free Snow Cover Products in China and Its High‐Mountain Region: Implications From Snow Cover Identification to Snow Phenology Detection

AbstractAccurate snow phenology detection, including snow cover days (SCD), snow start date (SSD), and snow end date (SED), is increasingly important for understanding mountain hydrology such as snow heterogeneity and snowmelt seasonality. Multiple cloud‐free daily snow cover products have recently been developed in China, employing diverse retrieval algorithms and cloud‐gap‐filling methods, resulting in varying accuracy levels. However, comprehensive analysis of differences among products and their impact on snow phenology detection is lacking. This study systematically evaluates eight state‐of‐the‐art snow cover products in China, focusing on the challenging Tibetan Plateau (TP). We introduce a novel metric, the consistency‐weighted correlation coefficient (CWR), customized for SSD and SED detection, and propose product‐combining schemes like “ensemble voting” and “sensor preference” to enhance reliability. Our findings highlight the prime influence of retrieval algorithms under clear‐sky conditions on accuracy, surpassing the importance of cloud‐gap‐filling methods. Specifically, a product optimizing normalized difference snow index thresholds for diverse landcover types consistently outperforms others in detecting all three snow phenology parameters, with correlation coefficients for SCD of 0.82 and 0.69, and CWR values for SSD of 0.54 and 0.40, and for SED of 0.53 and 0.37 in both China and the TP, respectively. Moreover, our proposed scheme combining three high‐accuracy products significantly enhances snow cover identification and SCD detection, especially when the best‐performing product alone faces substantial uncertainty. These findings provide immediate, crucial implications for optimizing the use of multiple cloud‐free products to enhance snow phenology detection, ultimately advancing the applicability of derived snow parameters in mountain hydrology research.

Predicting persistent fever in patients with cancer using machine learning.

e13552 Background: Sepsis is the leading cause of hospital readmission and death in the developed world. Cancer patients are at high risk due to their underlying disease and therapies. Meanwhile, antimicrobial resistance increases steadily, warranting a more careful administration of antibiotics. Predicting fever persistence 48 hours after initiation of antibiotics could help reduce antibiotic escalation and chest CT scans in patients where fever is predicted to subside within the next 24 hours. Methods: All cancer patients of the University Hospital Essen (UHE) between 1/1/2020-7/1/2023, receiving broad-spectrum i.v. antibiotics, and having persistent fever after 48 hours of first administration were extracted from UHE’s Fast Health Interoperability Resources clinical database. Vital signs, clinically important covariates, as well as lag features of vital signs and laboratory values were included. The binary target to predict was prevalence of body temperature ≥38 °C after 48-72 hours. Gradient boosting machines, a deep learning neural network, logistic regression (LR), random forest (RF) and a voting ensemble classifier were implemented. Hyperparameter tuning and model validation was performed through five-fold stratified cross validation (80/20% train/test split). Results: Analyzing 2.9 million patient records of UHE we obtained approx. 27,000 hospitalized adults with cancer. Restriction to hematology/oncology wards and intravenous broad-spectrum antibiotics resulted in 1,267 cancer patients. Of these, 264 patients had persistent fever after 48 hours, of which 111 patients had solid cancers, 66 lymphomas, and 87 other hematologic malignancies. 72 hours after antibiotic therapy 160 patients (60.2%) had persistent fever. Based on the area under the precision-recall curve (AUPRC) of 81.7%, the voting classifier model had the highest predictive performance. The model shows a sensitivity of 85%, a specificity of 63%, a Negative Predictive Value (NPV) of 75% and a Positive Predictive Value (PPV) of 73% on a cross-validated test set (Table). Conclusions: Machine learning-based algorithms can predict fever evolution and assist in clinical decision-making, potentially contributing to better patient care by preventing unnecessary antibiotics escalations and CT scans. Based on current model performance, 25% of chest CT scans could be obsolete as patients will not have persistent fever after 48 hours. Additional results will be presented at the meeting. [Table: see text]