FisherNet: AI-Driven Socio-Economic and Market Prediction for the Dry Fish Industry

In the realm of fisheries, particularly in the dry fish sector of South East Coast, Bangladesh, the twin challenges of predicting socio-economic outcomes for fishermen and forecasting market prices have significant implications.This study introduces a novel hybrid predictive model, dubbed FisherNet, designed to address these challenges by integrating a regression model for livelihood forecasting and a Seasonal Autoregressive Integrated Moving Average (SARIMA) model for price prediction.The foundation of this research is laid by a comprehensive survey encompassing 1657 participants from the dry fish industry in Cox's Bazar.The survey data, which includes occupational, personal, and production information, offers a detailed view of the current socio-economic status and market dynamics.This data undergoes rigorous descriptive and inferential statistical analysis, providing crucial insights into the living standards, work practices, and market strategies of the dry fish workers.FisherNet's architecture is a testament to the power of predictive modeling in addressing industry-specific challenges.The livelihood forecasting component of the model utilizes multiple regression analysis to predict socio-economic conditions, such as income levels and access to resources.Simultaneously, the SARIMA-based price forecasting model accurately predicts the market prices of dry fish, considering historical price data and seasonal variations.The integration of these two models in FisherNet is achieved through a sophisticated data fusion mechanism, providing a comprehensive outlook on how market trends might impact the socio-economic status of fishermen.The model boasts an impressive accuracy of 94.3%, with Mean Squared Error (MSE) of approximately 2417.27 and Root Mean Squared Error (RMSE) of about 49.17, indicating its robustness and reliability.