Abstract
Aquaculture is one of the most important activity in India that decides economic growth of the country. Currently, India ranked second in the world in terms of total fish production. In fish farming identifying the best fish species can grow in the fish pond is required that leads increase in production. In this based Deep Neural Networks (DNN) one of the deep learning models is used to identify the best fish species based on fish ponds water parameters like pH, temperature and turbidity. A dataset with 196 samples is developed by collecting real time data through pH sensor, temperature sensor and turbidity sensor in various ponds located at Warangal, Telangana State, India. The proposed DNN model is validated by comparing with other machine learning models like decision tree, random forest, SVC, KNN and Naive Bayes classifiers. Based on the simulation results it is observed that the proposed DNN model is performing better the other models with 100% accuracy in training and testing.