FCM Deep Neural Network

Fuzzy Cognitive Maps (FCMs), directed graphs representing causal relationships, are increasingly integrated with deep neural networks (DNNs) to model complex systems and improve prediction accuracy. Current research focuses on enhancing FCM extraction from data using techniques like large language models and developing robust FCM-DNN architectures, particularly for federated learning scenarios where data privacy is paramount. These advancements are impacting diverse fields, including healthcare (e.g., cancer detection, coronary artery disease diagnosis) and finance (e.g., financial distress prediction), by enabling more accurate and reliable predictive models while addressing challenges like data scarcity and security.