Modeling of cellular automata and agent-based complex systems

The term 'complex systems' may sound terrifying whenever you come across it as it depicts an overall collective structure which indeed can live upto its name; but when you comprehend the system at its fundamental level by stripping to its simpler multiple- interacting individual parts, the insights it provides may be used to describe and understand different problems ranging from atomic particles to the economics of societies and evolution. The simple laws can be used to simulate the behaviors of disparate complex systems.;In this thesis, a brief study is done emulating few such complex systems through programming techniques like cellular automata and neural networks. The patterns of complex behavior obtained are also classified respectively along with the help of Conway's game of life; the working of an autonomous and self organizing organism is simulated in a program written to show the complex patterns formed by a virtual ant. Then an important aspect of competition and cooperation among these agents is shown through game theory and dilemmas which throws light on the essence of survival of complex systems. A formal study is also done on the uses of artificial neural networks as associative memories and pattern recognizers.