Prediction algorithm for detecting closed areas for indiscriminate anchovy fishing, Callao

Abstract

Indiscriminate anchoveta fishing in Peru threatens the sustainability of this species and the balance of the marine ecosystem. The study proposes a prediction algorithm developed in Python to identify closed areas, evaluating variables such as quantity of fish, water temperature and age of anchovetas. The CRISP-DM methodology guides the process, divided into phases of data understanding, preparation and modeling, followed by model evaluation and deployment. The results show that areas such as Chorrillos exceed the 20% threshold of juvenile catch, indicating the need for stricter protection measures. In addition, an upward trend in the total catch of anchovetas is observed, suggesting an increase in demand or changes in the marine ecosystem. The cumulative probability analysis reveals stability in the juvenile population, and the scatter plots highlight the relationship between geographic coordinates and catches, aiding in fisheries management. The overall conclusion underlines the urgency of sustainable management, the implementation of monitoring technologies and the promotion of responsible fishing practices to protect marine biodiversity and guarantee the economic and food viability of the Peruvian fishing sector.