Optimizing Initial Guesses for Nonlinear System Solvers Using Machine Learning: A Comparative Study of Classification Algorithms DOI Open Access

Iven Aabaah,

Japheth Kodua Wiredu,

Bakaweri Emmanuel Batowise

и другие.

International Journal of Computer Science and Engineering, Год журнала: 2024, Номер 11(12), С. 7 - 15

Опубликована: Дек. 30, 2024

This paper focuses on the problem of improving initial guesses provided to solvers nonlinear systems in terms enhancing both convergence efficiency and reliability.A novel approach for constructing confidence models is proposed based a Logistic Regression, Support Vector Machines (SVM), Random Forests, K-Nearest Neighbors (KNN) classification schemes.Experimental evaluation across diverse highlights Forests as most effective model with an average accuracy 81.69%, precisionof 83.23%, recallof 82.16%, F1 score 82.69% highest AUC equal 0.90.Backed up by broad metrics, above research inquiries mark ideal potential machine learning revolutionize data processing increasing solver adaptability, patterns economizing computations scientific engineering modalities.

Язык: Английский

Enhanced Grey Wolf Optimization (EGWO) and random forest based mechanism for intrusion detection in IoT networks DOI Creative Commons

Saad Said Alqahtany,

Asadullah Shaikh, Ali Alqazzaz

и другие.

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Янв. 14, 2025

Smart devices are enabled via the Internet of Things (IoT) and connected in an uninterrupted world. These pose a challenge to cybersecurity systems due attacks network communications. Such have continued threaten operation end-users. Therefore, Intrusion Detection Systems (IDS) remain one most used tools for maintaining such flaws against cyber-attacks. The dynamic multi-dimensional threat landscape IoT increases Traditional IDS. focus this paper aims find key features developing IDS that is reliable but also efficient terms computation. Enhanced Grey Wolf Optimization (EGWO) Feature Selection (FS) implemented. function EGWO remove unnecessary from datasets intrusion detection. To test new FS technique decide on optimal set based accuracy achieved feature taking filters, recent approach relies NF-ToN-IoT dataset. selected evaluated by using Random Forest (RF) algorithm combine multiple decision trees create accurate result. experimental outcomes procedures demonstrate capacity recommended classification methods determine Analysis results presents performs more effectively than other techniques with optimized (i.e., 23 out 43 features), high 99.93% improved convergence.

Язык: Английский

Процитировано

1
Enhancing security of mobile crowd sensing in unmanned aerial vehicle ecosystems DOI Creative Commons

Sara Sumaidaa,

Hamda Almenhali,

Mohammed Alazzani

и другие.

Frontiers in Communications and Networks, Год журнала: 2025, Номер 6

Опубликована: Фев. 28, 2025

The rapid expansion of mobile devices with enhanced sensing and computing capabilities has driven the growth crowd (MCS), enabling applications that collect large datasets from sources like smartphones smartwatches. However, this data aggregation raises substantial security privacy concerns, especially when MCS integrates unmanned aerial vehicles (UAVs), where potential risks are further amplified. This study identifies analyzes specific threats in UAV-based through framework confidentiality, integrity, availability (CIA) triad. We categorize vulnerabilities propose comprehensive countermeasures targeting hardware, software, communication models. Our findings outline strategic actionable to mitigate identified risks, thus ensuring integrity reliable functionality within systems. Additionally, we present a scenario involving mitigation suggested for recovery. work underscores critical need robust frameworks UAV-enhanced applications, offering holistic approach emerging threats.

Язык: Английский

Процитировано

0
Optimized modified single shot multibox detector with hybrid encryption algorithm for satellite image security and classification DOI
J. Jeno Jasmine,

S. Edwin Raja,

R. Muniraj

и другие.

Earth Science Informatics, Год журнала: 2025, Номер 18(3)

Опубликована: Март 1, 2025

Язык: Английский

Процитировано

0
A Review of Ant Colony Optimization for Solving 0-1 Knapsack and Traveling Salesman Problems DOI Creative Commons

Isamadeen A. Khalifa,

Sagvan Ali Saleh

Deleted Journal, Год журнала: 2025, Номер 3(2), С. 87 - 99

Опубликована: Март 10, 2025

Ant Colony Optimization (ACO) represents a widespread nature-based metaheuristic algorithm which solves combinatorial optimization problems effectively [1]. This research study examines ACO-based solutions for Traveling Salesman Problem (TSP) and 0-1 Knapsack (0-1 KP) are both identified as NP-hard problems. ACO successfully achieves near-optimal because it duplicates real ants' pheromone-based foraging approach operates between exploration exploitation modes effectively. review discusses methods solving complex through discussion of modern solution their evaluation results performance benefits over basic approaches. section presents challenges include computational complexity two additional hybrid models while exploring adaptive parameter adjustments well quantum-inspired optimizations [2]. The development aims at combining this with deep learning reinforcement approaches to boost its operational speed practical across dynamic contexts. findings suggest that remains promising technique vast potential large-scale in various domains [3].

Язык: Английский

Процитировано

0
A Novel Beagle Inspired Optimization Algorithm: Comprehensive Evaluation on Benchmarking Functions DOI Creative Commons

Samindar Vibhute,

Chetan S. Arage

Research Square (Research Square), Год журнала: 2025, Номер unknown

Опубликована: Март 25, 2025

Abstract This paper presents an evaluation of the novel Beagle-Inspired Optimization Algorithm (BIOA), inspired by scent detection and rabbit hunting strategies beagle dogs, such as detection, tracking, trail following, pattern recognition, continuous adaptation, persistent exhaustive search, escape retrieval. BIOA is compared with well-established algorithms, including Particle Swarm (PSO), Artificial Bee Colony (ABC), Ant (ACO), Cuckoo Search (CS), across a set benchmark functions, Sphere, Rosenbrock, Rastrigin, Griewank, Ackley, Levy, Schwefel functions. The results demonstrate BIOA's superior performance, achieving lowest mean fitness values best solutions most test cases. Its balanced exploration exploitation phases enable effective optimization. While excels in many instances, it requires further improvements computational efficiency, particularly for high-dimensional problems. Future research should focus on enhancing performance through advanced models, hybrid optimization techniques, real-world problem applications, thus broadening its practical impact solving complex tasks.

Язык: Английский

Процитировано

0
Enhanced Intrusion Detection System Using a Two‐Staged Feature Selection Method DOI Open Access
R. Lalduhsaka, Ajoy Kumar Khan

Security and Privacy, Год журнала: 2025, Номер 8(3)

Опубликована: Март 28, 2025

ABSTRACT Intrusion detection (ID) systems are essential tools for safeguarding networks against cyber‐attacks. With the increasing sophistication and frequency of these attacks, developing ID that both accurate efficient is crucial. However, high‐dimensional datasets can hinder their efficiency increase computational costs. This paper proposes a novel two‐stage feature selection method (GIGA) to optimize enhance by reducing dimensionality while also improving accuracy. The first stage employs Gini impurity (GI) filter out features with less importance, followed Genetic Algorithm (GA) decision‐tree‐based fitness function identify most relevant subset features. Experiments on CIC‐IDS2017, CSE‐CIC‐IDS2018, CIC‐DDoS2019 demonstrate notable improvements: test accuracy increases from 99.31% 99.52%, 96.01% 97.19%, 97.95% 99.98%, respectively, False Positive Rate (FPR) decreases 0.71% 0.53%, 3.88% 1.04%, 0.03% 0.01%. number significantly reduced 71 8, 70 4, 69 8 datasets, respectively. proposed improves across machine learning models like Random Forest Decision Tree false positives negatives. By addressing key challenges in performance, GIGA offers scalable robust solution enhancing systems.

Язык: Английский

Процитировано

0
Improved deep learning model for accurate energy demand prediction and conservation in electric vehicles integrated with cognitive radio networks DOI Creative Commons

V Niranjani,

Anandakumar Haldorai

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 4, 2025

Язык: Английский

Процитировано

0
GINSER: Geographic Information System Based Optimal Route Recommendation via Optimized Faster R-CNN DOI Creative Commons

S.D. Anitha Selvasofia,

B. Sivasankari,

R. Dinesh

и другие.

International Journal of Computational Intelligence Systems, Год журнала: 2025, Номер 18(1)

Опубликована: Апрель 7, 2025

Язык: Английский

Процитировано

0
An Integrated Optimization Method for Resource-Constrained Schedule Compression Under Uncertainty in Construction Projects DOI Creative Commons

Firas Takleef,

Omar Ayadi, Faouzi Masmoudi

и другие.

Applied Sciences, Год журнала: 2025, Номер 15(8), С. 4089 - 4089

Опубликована: Апрель 8, 2025

An integrated solution that considers the shortening of scheduling and planning resource integration was conceived. The proposed method allocates resources execution mode costs effectively in order to minimize project duration cost construction activities. Costs are managed based on management already place for people those involved modes project, trying decrease as much possible. is used achieve maximum potential minimum during a including direct costs, indirect delay penalties. Furthermore, it finds balance between acquiring releasing human resources. most interesting aspect suggests addressing problems with simultaneously under uncertainty. FS theory model activity uncertainty method. In addition, above approach involves genetic algorithm (GA) schedule optimization. optimization utilizes GA an identify set non-dominated solutions. this paper, we discuss how string-based multi-object can be solved ES using elitist sorting (NSGA-II). implemented Python (v3.12.9), computer programming language, standalone automated computational tool subsequently reschedule.

Язык: Английский

Процитировано

0
Statistical optimisation of operational parameters using response surface methodology and phytoremediation of arsenic in constructed wetland DOI
Md Ekhlasur Rahman,

S. M. Shamsuzzaman,

Khairil Mahmud

и другие.

Chemistry and Ecology, Год журнала: 2025, Номер unknown, С. 1 - 31

Опубликована: Апрель 12, 2025

Язык: Английский

Процитировано

0