INTRODUCING THE ICN MODEL: A NOVEL APPROACH FOR PREDICTING CRITICAL POINTS IN SOCIAL TRENDS

In this paper, we present the ICN (Intersection Convergence Network) model as a novel approach for identifying and predicting critical points in social trend data. Unlike traditional models that focus on the overall trend fitting, the ICN model detects pivotal changes by analyzing the convergence of first and second derivatives. We apply the model to Brazil's homicide rate and GDP per capita data from ۱۹۹۰ to ۲۰۲۰, and compare its predictive performance with that of ARIMA and neural networks. The findings demonstrate that ICN provides improved accuracy and robustness, particularly in the presence of structural shifts and high data volatility.

کلیدواژه ها:

Critical Points Prediction ، Structural Change Recognition ، Social Trend Analysis ، Intersection Convergence Network (ICN)