Simultaneous discovery of multiple alternative optimal policies by reinforcement learning

Conventional reinforcement learning algorithms for direct policy search are limited to finding only a single optimal policy. This is caused by their local-search nature, which allows them to converge only to a single local optimum in policy space, and makes them heavily dependent on the policy initialization. In this paper, we propose a novel reinforcement learning algorithm for direct policy search, which is capable of simultaneously finding multiple alternative optimal policies. The algorithm is based on particle filtering and performs global search in policy space, therefore eliminating the dependency on the policy initialization, and having the ability to find the globally optimal policy. We validate the approach on one- and two-dimensional problems with multiple optima, and compare its performance to a global random sampling method, and a state-of-the-art Expectation-Maximization based reinforcement learning algorithm.