USING NEURAL NETWORK FOR TRAWL MANAGEMENT

The article highlights the problems of managing trawl fishing, increasing operation efficiency and reducing influence human factor. There has been considered using 
 a neural network in combination with mathematical model BigData technologies for predictive modeling process automatic control fishing order to increase its (to reduce energy labor costs, productivity). Advantages proposed approach are possibility account above factors neglected due complexity their description (e.g. time day, year, weather conditions, density, congestion, availability distribution food resources, other aquatic species), as well collecting accumulating data obtained many operations from different fishers subsequent consideration fishery management future. solution based on corrected output network. weight coefficients extracted centralized database before selection criterion area object fishing. In course input final (adjusted) recorded. At end saved is used training network, followed by updating database. learning occurs between fisheries shared adjusted updated general fishers.