Finding the high probabilistic potential fishing zone by accelerated SVM classification. (2017)

• Record Type: Journal Article
• Authors: Andrews, S.; Varunbabu, B.S.; Subash, P.; Swaminathan, M.R.
• Is Part Of: International journal of information and communication technology. Volume 11:Number 4(2017)
• Page Start: 576
• Page End: 585
• Publication Date: 2017
• Subjects: potential fishing zone -- PFZ -- fishing advisories -- Fourier transform -- support vector machines -- SVMs; Information technology -- Periodicals; Computer science -- Periodicals; Telecommunication -- Periodicals; 004.05
• Languages: English
• ISSNs: 1466-6642

Abstract

Potential fishing zones (PFZs) advisories play an important role in forecasting the spots for précised and optimised fishing. The PFZ advisories obtained from Indian National Centre for Ocean Information Services (INCOIS), during November 2003 to December 2011 of Kilakarai coast was analysed to understand the effective and high probabilistic fishing spot location. The original data for Kilakarai had four directions; they are south, southeast, southwest and east. INCOIS data is converted into FFT features to classify the areas using support vector machine (SVM) based on directions and plot the support vectors to distinguish locations. The identified support vectors for given direction combinations namely, south-southeast, south-southwest, south-east, southeast-southwest, southeast-east and southwest-east helps to a greater extent for classification of direction zone. SVM, a clustering technique, used to plot the support vectors to differentiate direction zones. We have accelerated the process by fine tuning the box constraint value which is used to plot the support vectors. In this work the high probabilistic potential fishing zones are effectively cropped to get the assured fishing zones.