An Automated Fish Species Classification Using Modified Neural Network and Firefly Algorithm and its Empirical Analysis

Oceans have always been a mystery and underwater environment had remained a topic of research and explorations. Fish species classification has gained huge attention among the researchers for obtaining a deeper insight into the marine ecological-system. This paper presents the simulation of a modified-Neural Network (NN) and Firefly Algorithm (FA) for classification of fish species of five different fish species using two algorithms Patternnet and CNN. The preliminary objective of this paper is to validate the potential of the two algorithms by analysing the performance with respect to diverse performance assessment metrics. Simulation results show that Patternnet exhibits better performance in terms of exactness as compared to CNN which shows good results in terms of execution time.