Evolutionary computation for feature selection in classification: A comprehensive survey of solutions, applications and challenges

Swarm and Evolutionary Computation, Год журнала: 2024, Номер 90, С. 101661 - 101661

Опубликована: Июль 22, 2024

Язык: Английский

---

Knowledge Assisted Differential Evolution Extreme Gradient Boost algorithm for estimating mangrove aboveground biomass

Applied Soft Computing, Год журнала: 2025, Номер unknown, С. 112838 - 112838

Опубликована: Фев. 1, 2025

Язык: Английский

---

Seasonal Land Use and Land Cover Mapping in South American Agricultural Watersheds Using Multisource Remote Sensing: The Case of Cuenca Laguna Merín, Uruguay

Sensors, Год журнала: 2025, Номер 25(1), С. 228 - 228

Опубликована: Янв. 3, 2025

Recent advancements in Earth Observation sensors, improved accessibility to imagery and the development of corresponding processing tools have significantly empowered researchers extract insights from Multisource Remote Sensing. This study aims use these technologies for mapping summer winter Land Use/Land Cover features Cuenca de la Laguna Merín, Uruguay, while comparing performance Random Forests, Support Vector Machines, Gradient-Boosting Tree classifiers. The materials include Sentinel-2, Sentinel-1 Shuttle Radar Topography Mission imagery, Google Engine, training validation datasets quoted methods involve creating a multisource database, conducting feature importance analysis, developing models, supervised classification performing accuracy assessments. Results indicate low significance microwave inputs relative optical features. Short-wave infrared bands transformations such as Normalised Vegetation Index, Surface Water Index Enhanced demonstrate highest importance. Accuracy assessments that various classes is optimal, particularly rice paddies, which play vital role country’s economy highlight significant environmental concerns. However, challenges persist reducing confusion between classes, regarding natural vegetation versus seasonally flooded vegetation, well post-agricultural fields/bare land herbaceous areas. Forests Trees exhibited superior compared Machines. Future research should explore approaches Deep Learning pixel-based object-based integration address identified challenges. These initiatives consider data combinations, including additional indices texture metrics derived Grey-Level Co-Occurrence Matrix.

Язык: Английский

---

Ensemble of neighborhood search operators for decomposition-based multi-objective evolutionary optimization

Expert Systems with Applications, Год журнала: 2025, Номер unknown, С. 127227 - 127227

Опубликована: Март 1, 2025

Язык: Английский

---

Differential Evolution Deep Reinforcement Learning Algorithm for Dynamic Multiship Collision Avoidance with COLREGs Compliance

Journal of Marine Science and Engineering, Год журнала: 2025, Номер 13(3), С. 596 - 596

Опубликована: Март 17, 2025

In ship navigation, determining a safe and economic path from start to destination under dynamic complex environment is essential, but the traditional algorithms of current research are inefficient. Therefore, novel differential evolution deep reinforcement learning algorithm (DEDRL) proposed address problems, which composed local planning global planning. The Deep Q-Network utilized search best in target multiple-obstacles scenarios. Furthermore, course-punishing reward mechanism introduced optimize constrain detected length as short possible. Quaternion domain COLREGs involved construct collision risk detection model. Compared with other algorithms, experimental results demonstrate that DEDRL achieved 28.4539 n miles, also performed all scenarios Overall, reliable robust for it provides an efficient solution avoidance.

Язык: Английский

---

Evolutionary Reinforcement Learning: A Systematic Review and Future Directions

Mathematics, Год журнала: 2025, Номер 13(5), С. 833 - 833

Опубликована: Март 2, 2025

In response to the limitations of reinforcement learning and Evolutionary Algorithms (EAs) in complex problem-solving, Reinforcement Learning (EvoRL) has emerged as a synergistic solution. This systematic review aims provide comprehensive analysis EvoRL, examining symbiotic relationship between EAs algorithms identifying critical gaps relevant application tasks. The begins by outlining technological foundations detailing complementary address learning, such parameter sensitivity, sparse rewards, its susceptibility local optima. We then delve into challenges faced both exploring utility EvoRL. EvoRL itself is constrained sampling efficiency algorithmic complexity, which affect areas like robotic control large-scale industrial settings. Furthermore, we significant open issues field, adversarial robustness, fairness, ethical considerations. Finally, propose future directions for emphasizing research avenues that strive enhance self-adaptation, self-improvement, scalability, interpretability, so on. To quantify current state, analyzed about 100 studies, categorizing them based on algorithms, performance metrics, benchmark Serving resource researchers practitioners, this provides insights state offers guide advancing capabilities ever-evolving landscape artificial intelligence.

Язык: Английский

---

Reinforcement learning for mutation operator selection in automated program repair

Automated Software Engineering, Год журнала: 2025, Номер 32(2)

Опубликована: Март 15, 2025

Abstract Automated program repair techniques aim to aid software developers with the challenging task of fixing bugs. In heuristic-based repair, a search space mutated variants is explored find potential patches for Most commonly, every selection mutation operator during performed uniformly at random, which can generate many buggy, even uncompilable programs. Our goal reduce generation that do not compile or break intended functionality waste considerable resources. this paper, we investigate feasibility reinforcement learning-based approach operators in repair. proposed programming language, granularity-level, and strategy agnostic allows easy augmentation into existing tools. We conducted an extensive empirical evaluation four techniques, two reward types, credit assignment strategies, integration methods, three sets using 30,080 independent attempts. evaluated our on 353 real-world bugs from Defects4J benchmark. The results higher number test-passing variants, but does exhibit noticeable improvement patched comparison baseline, uniform random selection. While learning has been previously shown be successful improving evolutionary algorithms, often used it yet demonstrate such improvements when applied area research.

Язык: Английский

---

A binary linear predictive evolutionary algorithm with feature analysis for multiobjective feature selection in classification

Engineering Applications of Artificial Intelligence, Год журнала: 2025, Номер 152, С. 110733 - 110733

Опубликована: Апрель 12, 2025

Язык: Английский

---

Logic mining method via hybrid discrete hopfield neural network

Computers & Industrial Engineering, Год журнала: 2025, Номер unknown, С. 111200 - 111200

Опубликована: Май 1, 2025

Язык: Английский

---

Reinforcement Intelligence for Spectral Enhancement (RISE): A Novel Feature Extraction Method for Hyperspectral Prediction of Sugar Content in Citrus reticulata 'Chun Jian'

Journal of Food Composition and Analysis, Год журнала: 2025, Номер unknown, С. 107857 - 107857

Опубликована: Май 1, 2025

Язык: Английский

---