Dimension reduction for microarray data using multi-objective ant colony optimisation

Ahuja, Jyoti; Ratnoo, Saroj

doi:10.1504/ijcsyse.2017.083149

Cited by 2 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Various traditional feature selection methods exist in the literature based on multi-objective optimization. DWFS: A wrapper feature selection tool based on a parallel genetic algorithm (Soufan et al, 2015) and Dimension reduction for microarray data using multi-objective ant colony optimization (Ahuja & Ratnoo, 2017) have been used as benchmark techniques to discover the performance of the novel proposed method. Soufan et al (2015) have implemented the wrapper model and applied parallel GA which simultaneously evaluates massive features.…”

Section: Benchmark Techniquesmentioning

confidence: 99%

“…DWFS incorporate diverse filtering methods and is used as a pre-processing step in feature selection (Soufan et al, 2015). Ahuja and Ratnoo et al (2017) have designed a multi-objective ant colony optimization (MOACO) algorithm to select genes in which several non-dominated solutions are selected instead of a single solution (Ahuja & Ratnoo, 2017). Here, a broad comparison of the proposed algorithm PMOGA has been made with these two previous approaches i.e.…”

Section: Benchmark Techniquesmentioning

confidence: 99%

See 1 more Smart Citation

Feature Selection using PMOGA for Microarray Datasets

Rathee¹,

Ratnoo²,

Ahuja³

2022

JSR

Self Cite

View full text Add to dashboard Cite

Microarray technology is significantly impacting the community to know the primary characteristic underlining the expansion and growth of genes. Regardless of its many helpful relevance in analysis of drug detection and disease diagnosis; microarray data has turned into a dispute for various bio-analysts. The dimensionality problem in Microarrays leads to expansion of some new methods. The dilemma of dimensionality reduction in terms of features has been taken as a multi-objective optimization problem, thus, can be solved by using some multi-objective optimization techniques.Yet reduction in features takes a lot of time for final submission, therefore, parallel genetic algorithms will do this task in a more efficient way by parallel optimization of multiple distinct parts of a dataset. In this paper, we have designed a new combined method for parallel implementation of gene selection in multi-objective perspective named as PMOGA. Individual migration strategy is followed to improve the parallel searching speed for improving the efficiency of the proposed algorithm. A comparative study of the proposed PMOGA has been done on eight most referenced datasets. The obtained results confirm the supremacy of MOGA based parallel approach over the other approaches based on different performance measures.

show abstract

Section: Benchmark Techniquesmentioning

confidence: 99%

Section: Benchmark Techniquesmentioning

confidence: 99%