scholarly journals Mathematical Morphology-Based Feature-Extraction Technique for Detection and Classification of Faults on Power Transmission Line

IEEE Access ◽  
2020 ◽  
Vol 8 ◽  
pp. 38459-38471 ◽  
Author(s):  
Revati Godse ◽  
Sunil Bhat
Keyword(s):  
Feature Extraction ◽  
Transmission Line ◽  
Mathematical Morphology ◽  
Power Transmission ◽  
Extraction Technique ◽  
Power Transmission Line ◽  
Feature Extraction Technique

A Novel Feature Extraction from Genome Sequences For Taxonomic Classification Of Living Organisms

2021 ◽  
Vol 12 (2) ◽  
pp. 1436-1451
Author(s):  
Muthulakshmi M, Et. al.
Keyword(s):  
Feature Extraction ◽  
Genome Sequence ◽  
Taxonomic Classification ◽  
Extraction Technique ◽  
Genome Sequences ◽  
National Centre ◽  
History Of ◽  
Living Organisms ◽  
Feature Extraction Technique

Genome sequencing aids in understanding the nature, characteristics, habitat and evolutionary history of all living organisms. Apart from sequencing, the more important task is to correctly place the sequenced genome in the taxonomy. Generally, the taxonomic classification of the living organisms is done by observing their morphological, behavioral, genetic and biochemical characteristics. Among them, taxonomic classification using genetic observation is more accurate scientifically as the Genome sequence analysis exploits the complete characteristics of the organism. In this paper, we developed a novel Frequency based Feature Extraction Technique (FFET) which extracts 120 features and helps to analyze the genome sequence of the organism and to classify them in the taxonomy accordingly. We performed a kingdom level taxonomic classification using the proposed FFET. The proposed FFET extracts features based on storage, frequency of nucleotide bases, pattern arrangement and amino acid composition of genome sequences. The feature extraction technique is applied to 150 samples of genome sequences of various organisms which were downloaded from National Centre for Biotechnology and Information (NCBI) database. The extracted features are classified using various Machine learning and Deep learning classifiers. From the results, it is evident that FFET performs well for classification with Convolutional Neural Network (CNN) classifier with an accuracy of 96.73 %.

Sign in / Sign up

Export Citation Format

Share Document