Optimized Feature Selection and Machine Learning for Accurate Celiac Disease Diagnosis with Model Interpretability

Deleted Journal, Journal Year: 2025, Volume and Issue: unknown

Published: April 11, 2025

Language: Английский

---

Enhanced Feature Selection Using Quantum-Inspired Cuckoo Search and Machine Learning for Heart Disease Prediction

IGI Global eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 339 - 372

Published: May 2, 2025

Heart disease remains a leading global health challenge demanding accurate predictive models for early diagnosis. Traditional machine learning (ML) struggle with high-dimensional data, feature selection, and interpretability in clinical settings. To address these challenges, we propose Quantum-Inspired Cuckoo Search Feature Selection Algorithm (QICSFA) integrating quantum principles optimized selection. Experimental results show that QICSFA combined Bayesian Optimization (BO) achieves 97% accuracy XGB 96% RF by outclassing conventional methods. The key features such as maximum heart rate (Thalach), chest pain type (Cp), ST depression (Oldpeak) align known cardiovascular risk factors to ensure relevance. In the future, this study establishes scalable AI-driven diagnostic tool potential applications real-time patient monitoring, multi-institutional dataset validation, explainable AI (XAI) integration, enhancing trust adoption healthcare systems.

Language: Английский

---

Chaotic marine predator optimization algorithm for feature selection in schizophrenia classification using EEG signals

Cluster Computing, Journal Year: 2024, Volume and Issue: 27(8), P. 11277 - 11297

Published: May 22, 2024

Abstract Schizophrenia is a chronic mental illness that can negatively affect emotions, thoughts, social interaction, motor behavior, attention, and perception. Early diagnosis still challenging based on the disease’s symptoms. However, electroencephalography (EEG) signals yield incredibly detailed information about activities functions of brain. In this study, hybrid algorithm approach proposed to improve search performance marine predator (MPA) chaotic maps. For evaluating chaotic-based (CMPA), benchmark datasets are used. The results suggested variation method benchmarks show Sine Chaotic-based MPA (SCMPA) significantly outperforms other variants. was verified using public dataset consisting 14 subjects. Moreover, SCMPA essential for EEG electrode selection because it minimizes model complexity selects best representative features providing optimal solutions. extracted each subject were used in decision tree (DT), random forest (RF), extra (ET) methods. Performance measures showed successful at differentiating schizophrenia patients (SZ) from healthy controls (HC). end, demonstrated feature technique SCMPA, which research, performs better regard classification signals.

Language: Английский

---

A Hybrid Feature Selection and Ensemble Stacked Learning Models on Multi-Variant CVD Datasets for Effective Classification

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 87023 - 87038

Published: Jan. 1, 2024

Language: Английский

---

Meta-heuristic optimization algorithms based feature selection for joint moment prediction of sit-to-stand movement using machine learning algorithms

Computers in Biology and Medicine, Journal Year: 2024, Volume and Issue: 178, P. 108812 - 108812

Published: June 28, 2024

Language: Английский

---

A highly accurate prediction for heart failure disease: a new deep attentive model with guided feature ranking

Arabian Journal for Science and Engineering, Journal Year: 2024, Volume and Issue: 49(9), P. 12167 - 12201

Published: Jan. 5, 2024

Abstract Heart failure (HF) is a life-threatening disease affecting at least 64 million people worldwide. Hence, it places great stresses on patients and healthcare systems. Accordingly, providing computerized model for HF prediction will help in enhancing diagnosis, treatment, long-term management of HF. In this paper, we introduce new guided attentive approach. method, sparse-guided feature ranking method proposed. Firstly, Gauss–Seidel strategy applied to the preprocessed pool low-rank approximation procedure with trace-norm regularization. The resultant sparse attributes, after Spearman elimination, are employed guide original through linear translation-variant model. Then, fast Newton-based non-negative matrix factorization pool. bases process finally utilized adopted deep predictive For final stage, instead commonly used machine learning approaches, an attentive-based classifier. It employs sequential attention choose most proper salient features efficient interpretability process. evaluation proposed model, three different datasets employed, i.e., UCI, Faisalabad, Framingham datasets. Compared state-of-the-art techniques, approach outperforms their performance all even small sizes. With only four bases, achieves average accuracy 98%, while, full gained.

Language: Английский

---

Fine-Tuned Cardiovascular Risk Assessment: Locally Weighted Salp Swarm Algorithm in Global Optimization

Mathematics, Journal Year: 2024, Volume and Issue: 12(2), P. 243 - 243

Published: Jan. 11, 2024

The Salp Swarm Algorithm (SSA) is a bio-inspired metaheuristic optimization technique that mimics the collective behavior of chains hunting for food in ocean. While it demonstrates competitive performance on benchmark problems, SSA faces challenges with slow convergence and getting trapped local optima like many population-based algorithms. To address these limitations, this study proposes locally weighted (LWSSA), which combines two mechanisms into standard framework. First, approach introduced integrated to guide search toward promising regions. This heuristic iteratively probes high-quality solutions neighborhood refines current position. Second, mutation operator generates new positions followers increase randomness throughout search. In order assess its effectiveness, proposed was evaluated against state-of-the-art metaheuristics using test functions from IEEE CEC 2021 2017 competitions. methodology also applied risk assessment cardiovascular disease (CVD). Seven strategies extreme gradient boosting (XGBoost) classifier are compared LWSSA-XGBoost model. achieves superior prediction 94% F1 score, recall, 93% accuracy, area under ROC curve comparison competitors. Overall, experimental results demonstrate LWSSA enhances SSA’s ability XGBoost predictive power automated CVD assessment.

Language: Английский

---

Advancing feature ranking with hybrid feature ranking weighted majority model: a weighted majority voting strategy enhanced by the Harris hawks optimizer

Journal of Computational Design and Engineering, Journal Year: 2024, Volume and Issue: 11(3), P. 308 - 325

Published: May 1, 2024

Abstract Feature selection (FS) is vital in improving the performance of machine learning (ML) algorithms. Despite its importance, identifying most important features remains challenging, highlighting need for advanced optimization techniques. In this study, we propose a novel hybrid feature ranking technique called Hybrid Ranking Weighted Majority Model (HFRWM2). HFRWM2 combines ML models with Harris Hawks Optimizer (HHO) metaheuristic. HHO known versatility addressing various challenges, thanks to ability handle continuous, discrete, and combinatorial problems. It achieves balance between exploration exploitation by mimicking cooperative hunting behavior Harris’s hawks, thus thoroughly exploring search space converging toward optimal solutions. Our approach operates two phases. First, an odd number models, conjunction HHO, generate encodings along metrics. These are then weighted based on their metrics vertically aggregated. This process produces rankings, facilitating extraction top-K features. The motivation behind our research 2-fold: enhance precision algorithms through optimized FS improve overall efficiency predictive models. To evaluate effectiveness HFRWM2, conducted rigorous tests datasets: “Australian” “Fertility.” findings demonstrate navigating We compared 12 other techniques found it outperform them. superiority was particularly evident graphical comparison dataset, where showed significant advancements ranking.

Language: Английский

---

Machine Learning-Based Surrogate Model for Genetic Algorithm with Aggressive Mutation for Feature Selection

International Journal of Hybrid Intelligent Systems, Journal Year: 2024, Volume and Issue: 20(3), P. 259 - 274

Published: July 16, 2024

The genetic algorithm with aggressive mutations GAAM, is a specialised for feature selection. This dedicated to the selection of small number features and allows user specify maximum desired. A major obstacle use this its high computational cost, which increases significantly dimensions be retained. To solve problem, we introduce surrogate model based on machine learning, reduces evaluations fitness function by an average 48% datasets tested, using standard parameters specified in original paper. Additionally, experimentally demonstrate that eliminating crossover step does not result any visible changes algorithm’s results. We also uses artificially complex mutation method could replaced simpler without loss efficiency. sum improvements resulted reduction 53% functions. Finally, have shown these outcomes apply beyond those utilized initial article, while still achieving comparable decrease count evaluation calls. Tests were conducted 9 varying dimensions, two different classifiers.

Language: Английский

---

Leveraging AI and Machine Learning for Next-Generation Clinical Decision Support Systems (CDSS)

Advances in healthcare information systems and administration book series, Journal Year: 2024, Volume and Issue: unknown, P. 83 - 112

Published: Nov. 27, 2024

Missed diagnoses and medication errors are significant risks in healthcare, leading to increased patient morbidity mortality. Traditional Clinical Decision Support Systems (CDSS) rely on static, predefined rules, limiting their adaptability personalized care. This chapter explores how integrating Artificial Intelligence (AI) Machine Learning (ML) can revolutionize CDSS, driving next-generation systems. By analyzing clinical datasets real time, AI ML enable insights that enhance diagnostic accuracy, optimize treatment recommendations, improve risk stratification, streamline workflows. These advancements promise better outcomes, informed decisions, reduced costs. The also addresses challenges like data quality, explainability, regulatory compliance, ethics, proposing strategies for overcoming these. Through collaboration research, transform CDSS into foundational healthcare elements, fostering personalized, data-driven, efficient

Language: Английский

---