GGSA: A Grouping Gravitational Search Algorithm for data clustering

Dowlatshahi, Mohammad Bagher; Nezamabadi‐pour, Hossein

doi:10.1016/j.engappai.2014.07.016

Cited by 78 publications

(21 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For any dataset, the best results are indicated in bold type. The reason is that the number of classi ers considered for comparison using three tness functions in the tables is not the same, which is due to the fact that there are no results on tness functions f 1 and f 3 in [30][31][32]. Table 9, EICA results are the best in comparison with the other algorithms based on tness function f 1 .…”

Section: Comparison Between Eica and Well-developed Algorithmsmentioning

confidence: 76%

“…First, in order to show the improvement obtained using EICA compared to the original ICA [33], we compare EICA with ICA. Second, we compare the performance of EICA with that of well-developed algorithms in classi cation (i.e., PSO [29], ABC [30], FA [31], GSA [28], and GGSA [32]). Third, the performance of EICA is compared against those of the other nine classi cation techniques well known in literature reported in [15] as: Bayes Net [55], MultiLayer Perception arti cial neural network (MLP) [56], Radial Basis Function arti cial neural network (RBF) [57], K Star [58], Bagging [59], Multi Boost AB [60], Naive Bayes Bayes Tree (NB Tree) [61], ripple down rule (Ridor) [62], and Voting Feature Interval (VFI) [63].…”

Section: Resultsmentioning

confidence: 99%

“…Bahrololoum et al [28] proposed a Gravitational Search Algorithm (GSA) to face the classi cation of instances in multi-class datasets. Recently, Dowlatshahi and Nezamabadi-pour [32] have presented a Grouping Gravitational Search Algorithm (GGSA) for data clustering. There are two signi cant advantages of GGSA algorithm.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A prototype-based classification using extended imperialist competitive algorithm

et al. 2017

View full text Add to dashboard Cite

KEYWORDSPrototype-based classi cation; Imperialist competitive algorithm; UCI machine learning repository.Abstract. Classi cation is an important machine learning technique used to predict group membership for data instances. In this paper, we propose an e cient prototypebased classi cation approach in the data classi cation literature by a novel soft-computing approach based on extended imperialist competitive algorithm. The novel classi er is called EICA. The goal is to determine the best places of the prototypes. EICA is evaluated under three di erent tness functions on twelve typical test datasets from the UCI Machine Learning Repository. The performance of the proposed EICA is compared with well-developed algorithms in classi cation including original Imperialist Competitive Algorithm (ICA), the Arti cial Bee Colony (ABC), the Fire y Algorithm (FA), the Particle Swarm Optimization (PSO), the Gravitational Search Algorithm (GSA), the Grouping Gravitational Search Algorithm (GGSA), and nine well-known classi cation techniques in the literature. The analysis results show that EICA provides encouraging results in contrast to other algorithms and classi cation techniques.

show abstract

Section: Comparison Between Eica and Well-developed Algorithmsmentioning

confidence: 76%

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

A prototype-based classification using extended imperialist competitive algorithm

et al. 2017

View full text Add to dashboard Cite

show abstract

“… Adapting the grouping representation, [16], we proposed a Grouping Gravitational Search Algorithm (GGSA) for data clustering in this paper.  The main property [16], of the grouping representation which encourages us to use it, is that it has very low redundancy also it have some grouping problems.…”

Section: Problem Identificationmentioning

confidence: 99%

“…The author, M. B. Dowlatshahi and H. Nezamabadi-pour [16] have developed the structure of GSA for solving the data clustering problem, the problem of grouping data into clusters such that the data in each cluster share a high degree of similarity while being very dissimilar to data from other clusters. The proposed algorithm, which was called Grouping GSA (GGSA), differs from the standard GSA in two important aspects.…”

Section: Literature Surveymentioning

confidence: 99%

Optimal Decision Tree Based Unsupervised Learning Method for Data Clustering

Seelam¹,

Reddy²,

Mukkala³

2017

IJIES

View full text Add to dashboard Cite

Abstract:Clustering is an investigative data analysis task. It aims to find the intrinsic structure of data by organizing data objects into similarity groups or clusters. Our investigation using a pattern based clustering on numerical data set; here, we are using a Parkinson and spam dataset. These techniques are strongly related to the statistical field of cluster analysis, where over the years a large number of clustering methods has been proposed. Here, we have proposed an improved k-means clustering algorithm is used to extract patterns from a collection of an unsupervised decision tree. In our proposed research, we introduce a binary cuckoo search based decision tree. In this tree based learning technique, extracting patterns from a given dataset. Here, we have clustered the data with the aid of improved k-means clustering algorithm. The performance can be evaluated in terms of sensitivity, specificity, and accuracy.

show abstract

Ensemble feature selection for multi‐label text classification: An intelligent order statistics approach

Miri

Dowlatshahi

Hashemi

et al. 2022

Int J of Intelligent Sys

Self Cite

View full text Add to dashboard Cite

Because of the overgrowth of data, especially in text format, the value and importance of multi‐label text classification have increased. Aside from this, preprocessing and particularly intelligent feature selection (FS) are the most important step in classification. Each FS finds the best features based on its approach, but we try to use a multi‐strategy approach to find more useful features. Evaluating and comparing features’ importance and relevance makes using multiple strategy and methods more suitable than conventional approaches because each feature is measured based on several perspectives. Nevertheless, the ensemble FS merges the final performance results of various methods to take advantage of different methods’ strengths and better classify. In this article, we have proposed an ensemble FS method for multi‐label text data (MLTD) for the first time using the order statistics (EMFS) approach. We have utilized four multi‐label FS (MLFS) algorithms with various particular performances to achieve a good result. In this method, as one of the most important statistics methods, Order Statistics was used to aggregate the ranks of different algorithms, which is robust against noise, redundant and inessential features. In the end, the performance of EMFS, executing six MLTDs, was evaluated according to six performance criteria (ranking‐based and classification‐based). Surprisingly, the proposed method was more accurate than others among all used MLTDs. The proposed method has improved by 1.5% compared to other methods. This value is based on the results obtained based on six evaluation criteria and all tested data sets.

show abstract

GGSA: A Grouping Gravitational Search Algorithm for data clustering

Cited by 78 publications

References 35 publications

A prototype-based classification using extended imperialist competitive algorithm

A prototype-based classification using extended imperialist competitive algorithm

Optimal Decision Tree Based Unsupervised Learning Method for Data Clustering

Ensemble feature selection for multi‐label text classification: An intelligent order statistics approach

Contact Info

Product

Resources

About