Inverted Pendulum control using Bayesian networks

The project utilises the Bayesian approach for the problem of inverted pendulum control. The method alleviates the planning in action space encountered in Markov Decision Process. The algorithm for the control of inverted pendulum builds upon the Markov Decision Process and models the control parameters with Bayesian Networks. The state space parameters of the pole position are further implemented using a binary tree data structure to reduce the running time complexity of MDP based algorithms.