Reinforcement Learning and Dynamic Programming Techniques - Research Paper Example

Both pervasive directional graphs and data intensive algebraic operations require adaptive search functions in problem solving and reasoning with limited computing resources. Among search methods, algorithmic complexity, such as memory bound problems, error convergence issues and supervised training are prohibitive for large state and solution spaces or high dimensional state spaces. In addition, among popular search strategies, heuristic algorithms many1 not guarantee search optimality or hard to approximate without close formed utility functions. Furthermore, model-building algorithms require large computation per iteration since every update needs to compute sums over the entire state space. Hence, dynamic search function and control optimization are major primitives to construct search utilities for stochastic system processes to ensure converged resource accesses. This research focuses on optimization of general search solution methods and proposes a formal search utility framework, algorithms rooted from Reinforcement Learning (RL) and Dynamic Programming (DP) techniques. To reduce space complexity within large dimension search spaces, a memoryless2 Q learning is augmented with self-organized index structure and algorithms for exact state-action value function mapping to optimize search procedures for optimal policies. Data parallelization is ensured with this paged based index value mapping function. Hence, time complexity is reduced with threaded search parallelism. Convergence analysis and error estimation are presented for numeric and information evaluation. Finally, simulation and learning results are presented and discussed. For search strategies in the settings of problem solving and reasoning, search problem formulation represents many combinatorial optimization problems of search approximation with action control optimization. All aspects of search task environment represent various classes of applications, such as routing, scheduling, speech recognition, scene analysis and intrusion detection pattern matching. By given a directional graph G with distinguished starting state Sstart and a set of goal states Read More