FUZZY COGNITIVE MAP AND NONLINEAR HEBBIAN LEARNING ALGORITHMS FOR MODELLING AND CONTROLLING INTRA-STATE CONFLICT IN NIGERIA

Abstract

Ethnic conflict, communal clashes, terrorism and insurgency are major security challenges in the twenty-first century. These are responsible for the deaths of millions of people around the world. At the moment intra-state conflict in the form of Farmer/herder conflict has claimed several thousands of lives in Nigeria; however, controlling this conflict system has been a great challenge for many years. Combining Fuzzy Cognitive Map (FCM) and Non-liner Hebbian Learning (NHL), this work modelled Farmer/Herder conflict in Nigeria as a nonlinear system and demonstrates the ability of the machine learning algorithms to provide control mechanism for the conflict system. Factors understood to be influential in Farmer/Herder conflict were used to form FCM model that represents the conflict situation, while NHL was used as a control mechanism to find levels of weights between these causative factors that will minimize the conflict and maximize peace. The result of the work showed that keeping certain factors within some threshold and ensuring these factors interact at particular levels will reduce conflict and bring peace.