Symbolic Network: Generalized Neural Policies for Relational MDPs
This addresses the challenge of scalable probabilistic planning for AI systems by enabling efficient, instance-independent policies, though it is incremental as it builds on prior neural and relational methods.
The paper tackles the problem of solving Relational Markov Decision Processes (RMDPs) with unbounded objects by proposing SymNet, a neural approach that trains shared parameters on domain instances to compute generalized policies, achieving results significantly better than random and sometimes outperforming state-of-the-art deep reactive policies in experiments on nine RDDL domains.
A Relational Markov Decision Process (RMDP) is a first-order representation to express all instances of a single probabilistic planning domain with possibly unbounded number of objects. Early work in RMDPs outputs generalized (instance-independent) first-order policies or value functions as a means to solve all instances of a domain at once. Unfortunately, this line of work met with limited success due to inherent limitations of the representation space used in such policies or value functions. Can neural models provide the missing link by easily representing more complex generalized policies, thus making them effective on all instances of a given domain? We present SymNet, the first neural approach for solving RMDPs that are expressed in the probabilistic planning language of RDDL. SymNet trains a set of shared parameters for an RDDL domain using training instances from that domain. For each instance, SymNet first converts it to an instance graph and then uses relational neural models to compute node embeddings. It then scores each ground action as a function over the first-order action symbols and node embeddings related to the action. Given a new test instance from the same domain, SymNet architecture with pre-trained parameters scores each ground action and chooses the best action. This can be accomplished in a single forward pass without any retraining on the test instance, thus implicitly representing a neural generalized policy for the whole domain. Our experiments on nine RDDL domains from IPPC demonstrate that SymNet policies are significantly better than random and sometimes even more effective than training a state-of-the-art deep reactive policy from scratch.