An improved soft voting-based machine learning technique to detect breast cancer utilizing effective feature selection and SMOTE-ENN class balancing

Discover Artificial Intelligence, Journal Year: 2025, Volume and Issue: 5(1)

Published: Jan. 20, 2025

Language: Английский

---

Synergistic Feature Engineering and Ensemble Learning for Early Chronic Disease Prediction

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 62215 - 62233

Published: Jan. 1, 2024

Chronic diseases, a global public health challenge, necessitate the deployment of cutting-edge predictive models for early diagnosis and personalized interventions. This study presents an advanced methodology prediction chronic including heart attack, diabetes, breast cancer, kidney disease, leveraging synergistic combination techniques. Recognizing challenge posed by extensive medical datasets with numerous features, we introduce novel approach that begins Feature Engineering using Recursive Elimination (RFE) in conjunction Support Vector Machine (SVM). The presented identifies removes irrelevant features to simplify data complexity. refined dataset is then input into robust eXtreme Gradient Boosting (XGBoost) classifier, known its efficiency adeptness predicting complex relationships within data. chosen ensemble learning algorithm demonstrates significant prowess inducing intricate patterns crucial disease prediction. To enhance model performance, essential phase optimization introduced through hyperparameter tuning Bayesian optimization. strategically navigates space, maximizing search process fine-tuning optimal accuracy. proposed showcases substantial improvement demonstrating effectiveness approach.

Language: Английский

---

Enhanced photovoltaic panel diagnostics through AI integration with experimental DC to DC Buck Boost converter implementation

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 2, 2025

Language: Английский

---

The role of machine learning algorithms in detection of gestational diabetes; a narrative review of current evidence

Clinical Diabetes and Endocrinology, Journal Year: 2024, Volume and Issue: 10(1)

Published: June 25, 2024

Abstract Gestational Diabetes Mellitus (GDM) poses significant health risks to mothers and infants. Early prediction effective management are crucial improving outcomes. Machine learning techniques have emerged as powerful tools for GDM prediction. This review compiles analyses the available studies highlight key findings trends in application of machine A comprehensive search relevant published between 2000 September 2023 was conducted. Fourteen were selected based on their focus These subjected rigorous analysis identify common themes trends. The revealed several themes. Models capable predicting risk during early stages pregnancy identified from reviewed. Several underscored necessity tailoring predictive models specific populations demographic groups. highlighted limitations uniform guidelines diverse populations. Moreover, emphasised value integrating clinical data into models. integration improved treatment care delivery individuals diagnosed with GDM. While different showed promise, selecting weighing variables remains complex. reviewed offer valuable insights complexities potential solutions using learning. pursuit accurate, models, consideration populations, data, emerging sources underscore commitment researchers improve healthcare outcomes pregnant at

Language: Английский

---

Machine Learning for Predicting and Optimizing Physicochemical Properties of Deep Eutectic Solvents: Review and Perspectives

Industrial & Engineering Chemistry Research, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 19, 2024

This review explores the application of machine learning in predicting and optimizing key physicochemical properties deep eutectic solvents, including CO2 solubility, density, electrical conductivity, heat capacity, melting temperature, surface tension, viscosity. By leveraging learning, researchers aim to enhance understanding customization a critical step expanding their use across various industrial research domains. The integration represents significant advancement tailoring solvents for specific applications, marking progress toward development greener more efficient processes. As continues unlock full potential it is expected play an increasingly pivotal role revolutionizing sustainable chemistry driving innovations environmental technology.

Language: Английский

---

Optimizing Predictive Performance: Hyperparameter Tuning in Stacked Multi-Kernel Support Vector Machine Random Forest Models for Diabetes Identification

Journal of Robotics and Control (JRC), Journal Year: 2024, Volume and Issue: 4(6), P. 896 - 904

Published: Jan. 5, 2024

This study addresses the necessity for more advanced diagnostic tools in managing diabetes, a chronic metabolic disorder that leads to disruptions glucose, lipid, and protein metabolism caused by insufficient insulin activity. The research investigates innovative application of machine learning models, specifically Stacked Multi-Kernel Support Vector Machines Random Forest (SMKSVM-RF), determine their effectiveness identifying complex patterns medical data. ensemble method SMKSVM-RF combines strengths (SVMs) Forests (RFs) leverage diversity complementary features. SVM component implements multiple kernels identify unique data patterns, while RF consists an decision trees ensure reliable predictions. Integrating these models into stacked architecture allows enhance overall predictive performance classification or regression tasks optimizing strengths. A significant finding this is introduction SMKSVM-RF, which displays impressive 73.37% accuracy rate confusion matrix. Additionally, its recall 71.62%, precision 70.13%, it has noteworthy F1-Score 71.34%. technique shows potential enhancing current methods developing ideal healthcare system, signifying step forward diabetes detection. results emphasize importance sophisticated methods, highlighting how can improve aid continual advancement systems effective management.

Language: Английский

---

Investigating Gender and Age Variability in Diabetes Prediction: A Multi-Model Ensemble Learning Approach

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 71535 - 71554

Published: Jan. 1, 2024

The study investigates the intricate influence of gender and age variability in individuals diagnosed with diabetes, aiming to gain a comprehensive understanding diverse impact implications this prevalent metabolic disorder. A real-world dataset, obtained from renowned diabetologist meticulously maintained by Dr. Reddys' Lab, serves as foundation for rigorous analysis. Leveraging capabilities ensemble learning, an advanced technique that combines multiple models, predictive model's efficiency is substantially enhanced, resulting precise reliable predictions individuals' diabetic status. Addressing challenge diabetes prediction, novel learning model was proposed. strengths three distinct algorithms: Random Forest, Extra Trees, Multilayer Perceptron (MLP). output comprises ternary label categorizing "diabetic, non-diabetic, or pre-diabetic", while accompanying prediction score quantifies likelihood belonging each respective category. findings research expand existing body knowledge on underscoring untapped potential methodologies augmenting accuracy performance patients.

Language: Английский

---

MACHINE LEARNING BASED PREDICTION AND INSIGHTS OF DIABETES DISEASE: PIMA INDIAN AND FRANKFURT DATASETS

JOURNAL OF MECHANICS OF CONTINUA AND MATHEMATICAL SCIENCES, Journal Year: 2025, Volume and Issue: 20(1)

Published: Jan. 18, 2025

This paper focused on predicting diabetes disease using machine learning models which is a very active and highly important area of research. Six methods three datasets were experimented with to investigate model performances. The are logistic regression, k-Nearest Neighbour, Gaussian Naïve Bayes, Decision Tree, Random Forest, XGBoost. Pima Indian, the Frankfurt Hospital dataset, combined dataset where all have 08 (eight) feature variables 01 (one) target variable. Train-test data split ratio can make significant difference in performance. Hence, two different ratios 50-50 90-10 experimented. Model performances evaluated four performance metrics precision, recall, F1-score, accuracy. Forest XGBoost found be efficient best-performing among based metrics, datasets, both train-test ratios. They performed comparatively better involved 2768 instances indicating importance large for results. Also, produced improved results than even almost models.

Language: Английский

---

Physiological, Psychological, and Functional Health Determinants of Depressive Symptoms Among the Elderly in India: Evaluation of Classification Performance of XGBoost Models

Indian Journal of Psychological Medicine, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 25, 2025

Depression among the elderly is a growing public health concern, especially in India. This study aimed to investigate predictive validity of physiological, psychological, and functional factors classifying level depressive symptoms using extreme gradient boosting (XGBoost) technique. Additionally, we compared performance models trained on original resampled data. entirely based secondary data analysis Longitudinal Aging Study India wave 1 We classified observations into "high symptom" "low/no groups predictors, including factors, along with socio-demographic factors. developed three (Models 1, 2, 3) original, over-sampled, under-sampled data, respectively. Model was evaluated metrics balanced accuracy, sensitivity, specificity, area under receiver operating characteristics curve (AUC). The included 26,065 individuals aged 60 above. 3, demonstrated best overall performance. It achieved accuracy 64%, sensitivity 62.8% specificity 65.2%. AUC for 3 0.692. Feature importance revealed that life satisfaction, instrumental activities daily living, mobility, caste, monthly per capita expenditure quintiles were most influential predicting symptoms. XGBoost demonstrate promise elderly. These findings suggest machine learning can be envisaged early detection management depression, primary care.

Language: Английский

---

Advanced Machine Learning Algorithms for Personalized Diabetic Foot Ulcer Treatment

IGI Global eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 347 - 372

Published: Feb. 5, 2025

Personalized treatment for diabetic foot ulcers (DFUs) is crucial optimal healing, minimizing complications, and reducing healthcare costs. Machine learning (ML) offers solutions to DFU complexities through accurate predictions, timely interventions, customized treatments. This chapter delves into advanced ML algorithms care. Ensemble techniques like random forests gradient boosting enhance model performance interpretability. Deep architectures such as recurrent neural networks (RNNs) convolutional (CNNs) process sequential data images. Reinforcement learning, including Q-learning, proposed adaptive plans from real-world inputs. Applications of these include predictive analytics, early detection systems, multi-factor decision-making. Case studies highlight their clinical effectiveness, improved patient outcomes, valuable insights. underscores ML's transformative role in advancing personalized treatment.

Language: Английский

---