A PSO based hybrid feature selection algorithm for high-dimensional classification

Tran, Binh; Zhang, Mengjie; Xue, Bing

doi:10.1109/cec.2016.7744271

Cited by 43 publications

(30 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Conversely, the proposed approach underperforms for -11 Tumors‖ datasets with merely 0.33% difference in terms of accuracy. Additionally, the proposed approach is 5-8 times faster than the approach suggested by authors [20,25].…”

Section: Comparison With Existing Methodsmentioning

confidence: 76%

“…Best results obtained in proposed work are compared with the best results of some existing work as shown in Table 5. As seen in Table 5 that the proposed approach selects the minimum number of features as compared to the methods proposed by authors [20,25]. TC-BPSO not only selects the minimum number of features but also improves the classification accuracy.…”

Section: Comparison With Existing Methodsmentioning

confidence: 90%

“…Tran et al [20] propose a new hybrid PSO-based approach (PSO-LSSU) where the evolutionary process of PSO combines wrapper method and filter method. PSO-LSSU obtains higher accuracy classification and lesser time for the feature selection of higher dimensional data.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Threshold Controlled Binary Particle Swarm Optimization for High Dimensional Feature Selection

Gupta¹,

Baghel²,

Iqbal³

2018

IJISA

View full text Add to dashboard Cite

Dimensionality reduction or the optimal selection of features is a challenging task due to large search space. Currently, many research has been performed in this domain to improve the accuracy as well as to minimize the computational complexity. Particle Swarm Optimization (PSO) based feature selection approach seems very promising and has been extensively used for this work. In this paper, a Threshold Controlled Binary Particle Swarm Optimization (TC-BPSO) along with Multi-Class Support Vector Machine (MC-SVM) is proposed and compared with Conventional Binary Particle Swarm Optimization (C-BPSO). TC-BPSO is used for the selection of features while MC-SVM is used to calculate the classification accuracy. 70% of the data is used to train the MC-SVM model while the test has been performed on rest 30% data to calculate the accuracy. Proposed approach is tested on ten different datasets having varying difficulties such as some datasets having large number of features while some have small, some have just two classes while some have many classes, some datasets having small number of instances while some have large number of instances and the results obtained on these datasets are compared with some of the existing methods. Experiments show that the obtained results are very promising and achieved the best accuracy in minimum possible features. Proposed approach outperforms C-BPSO in all contexts on most of the datasets and 3-4 times computationally faster. It also outperforms in all context when compared with the existing work and 5-8 times computationally faster.

show abstract

Section: Comparison With Existing Methodsmentioning

confidence: 76%

Section: Comparison With Existing Methodsmentioning

confidence: 90%

See 1 more Smart Citation

Threshold Controlled Binary Particle Swarm Optimization for High Dimensional Feature Selection

Gupta¹,

Baghel²,

Iqbal³

2018

IJISA

View full text Add to dashboard Cite

show abstract

“…Based on the evaluation criteria, feature selection algorithms are generally classified into two categories: 1) filter approaches and 2) wrapper approaches [Ambusaidi, He, Nanda et al (2016); Zawbaa, Emary, Hassanien et al (2016)]. Their main difference is that wrapper approaches include a classification/learning algorithm in feature subset evaluation, while filter algorithms are independent of any classification algorithms [Tran, Zhang and Xue (2016)]. Filters ignore the performance of the selected features, whereas wrappers not, which usually results in those wrappers getting better solutions [Yang, Liu, Zhou et al (2013)].…”

Section: Related Workmentioning

confidence: 99%

Feature Selection with a Local Search Strategy Based on the Forest Optimization Algorithm

Ma¹,

Zhou²,

Jia³

et al. 2019

Computer Modeling in Engineering &Amp; Sciences

View full text Add to dashboard Cite

Feature selection has been widely used in data mining and machine learning. Its objective is to select a minimal subset of features according to some reasonable criteria so as to solve the original task more quickly. In this article, a feature selection algorithm with local search strategy based on the forest optimization algorithm, namely FSLSFOA, is proposed. The novel local search strategy in local seeding process guarantees the quality of the feature subset in the forest. Next, the fitness function is improved, which not only considers the classification accuracy, but also considers the size of the feature subset. To avoid falling into local optimum, a novel global seeding method is attempted, which selects trees on the bottom of candidate set and gives the algorithm more diversities. Finally, FSLSFOA is compared with four feature selection methods to verify its effectiveness. Most of the results are superior to these comparative methods.

show abstract

“…To verify the performance of the proposed BBPSO-ACJ, the following 8 EC based wrappers are employed: Genetic algorithm (GA) 40 , PSO 31 , Binary PSO (BPSO) 30 ,Binary PSO with chaotic inertia weight (BPSO-CI) 35 , BBPSO 23 , Quantum inspired PSO (QBPSO) 41 , Binary PSO with catfish effect (BPSO-CE) 15 , PSO (4-2) 18 . Furthermore, two filter based methods, linear forward selection (LFS) and greedy stepwise based selection (GSBS), are also employed for comparison.…”

Section: Comparative Algorithmsmentioning

confidence: 99%

Bare bones particle swarm optimization with adaptive chaotic jump for feature selection in classification

Qiu¹

2018

IJCIS

View full text Add to dashboard Cite

Feature selection (FS) is a crucial data pre-processing process in classification problems. It aims to reduce the dimensionality of the problem by eliminating irrelevant or redundant features while achieve similar or even higher classification accuracy than using all the features. As a variant of particle swarm optimization (PSO), Bare bones particle swarm optimization (BBPSO) is a simple but very powerful optimizer. However, it also suffers from premature convergence like other PSO algorithms, especially in high-dimensional optimization problems. In order to improve its performance in FS problems, this paper proposes a novel BBPSO based FS method called BBPSO-ACJ. An adaptive chaotic jump strategy is designed to help the stagnated particles make a large change in their searching trajectory. It can enrich the search behavior of BBPSO and prevent the particles from being trapped into local attractors. A new global best updating mechanism is employed to reduce the size of obtained feature subset. The proposed BBPSO-ACJ is compared with eight evolutionary computation (EC) based wrapper methods and two filter methods on nine benchmark datasets with different number of dimensions and instances. The experimenta l results indicate that the proposed method can select the most discriminative features from the entire feature set and achieve significantly better classification performance than other comparative methods.

show abstract

A PSO based hybrid feature selection algorithm for high-dimensional classification

Cited by 43 publications

References 28 publications

Threshold Controlled Binary Particle Swarm Optimization for High Dimensional Feature Selection

Threshold Controlled Binary Particle Swarm Optimization for High Dimensional Feature Selection

Feature Selection with a Local Search Strategy Based on the Forest Optimization Algorithm

Bare bones particle swarm optimization with adaptive chaotic jump for feature selection in classification

Contact Info

Product

Resources

About