Dynamic evolutionary data and text document clustering approach using improved Aquila optimizer based arithmetic optimization algorithm and differential evolution

Data and text clustering are popular and frequently used in the data mining domain, mainly to deal with big data analytics. The main problem in these techniques is finding the most coherent clusters allocating similar-related objects into one group. In this paper, an improved clustering analysis app...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Neural computing & applications Ročník 34; číslo 23; s. 20939 - 20971
Hlavní autoři: Abualigah, Laith, Almotairi, Khaled H
Médium: Journal Article
Jazyk:angličtina
Vydáno: London Springer London 01.12.2022
Springer Nature B.V
Témata:
Algorithms
Arithmetic
Artificial Intelligence
Big Data
Cluster analysis
Clustering
Computational Biology/Bioinformatics
Computational Science and Engineering
Computer Science
Data mining
Data Mining and Knowledge Discovery
Datasets
Documents
Evolutionary computation
Image Processing and Computer Vision
Optimization
Optimization algorithms
Original Article
Performance evaluation
Probability and Statistics in Computer Science
Search methods
Standard data
Arithmetic optimization algorithm (AOA)
Text document clustering
Aquila optimizer (AO)
Data clustering
Differential evolution (DE)
ISSN:0941-0643, 1433-3058
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Data and text clustering are popular and frequently used in the data mining domain, mainly to deal with big data analytics. The main problem in these techniques is finding the most coherent clusters allocating similar-related objects into one group. In this paper, an improved clustering analysis approach is proposed using an advanced optimization method called AOAOA. The proposed AOAOA method improved the Aquila optimizer (AO) search performance by the operators of the arithmetic optimization algorithms (AOA) and differential evolution (DE) and using a novel transition mechanism. The primary motivation for this modification is that the original optimizer suffers from local optima stagnation and lacks search balance. Thus, the proposed AOAOA overcame these shortcomings by integrating various powerful search strategies and a new update strategy. Experiments are conducted on two parts; eight standard data clustering datasets and ten text documents benchmark datasets to evaluate the performance of the proposed AOAOA method. The proposed method is compared against several well-known optimization algorithms and advanced state-of-the-art methods published in the literature. The data clustering results also showed promising performance for the proposed AOAOA compared to other comparative data clustering methods. Moreover, the results illustrated that the proposed AOAOA can find new best solutions for several different complicated cases as the text document clustering results. The proposed AOAOA got accurate and robust results compared to several state-of-the-art methods.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0941-0643
1433-3058
DOI:10.1007/s00521-022-07571-0