Hill Climbing on Value Estimates for Search-control in Dyna
Dyna is an architecture for model-based reinforcement learning (RL), where simulated experience from a model is used to update policies or value functions. A key component of Dyna is search-control, the mechanism to generate the state and action from which the agent queries the model, which remains largely unexplored. In this work, we propose to generate such states by using the trajectory obtained from Hill Climbing (HC) the current estimate of the value function. This has the effect of propagating value from high-value regions and of preemptively updating value estimates of the regions that the agent is likely to visit next. We derive a noisy projected natural gradient algorithm for hill climbing, and highlight a connection to Langevin dynamics. We provide an empirical demonstration on four classical domains that our algorithm, HC-Dyna, can obtain significant sample efficiency improvements. We study the properties of different sampling distributions for search-control, and find that there appears to be a benefit specifically from using the samples generated by climbing on current value estimates from low-value to high-value region.
Code (0)
등록된 구현이 없습니다.
Tasks
Model-based Reinforcement LearningReinforcement LearningReinforcement Learning (RL)Similar Papers 제목 키워드 기반
Bandit-Based Random Mutation Hill-Climbing
The Random Mutation Hill-Climbing algorithm is a direct search technique mostly used in discrete domains. It repeats the process of randomly selecting a neighbour of a best-so-far solution and accepts the neighbour if it…
Hybrid Genetic Algorithm and Hill Climbing Optimization for the Neural Network
In this paper, we propose a hybrid model combining genetic algorithm and hill climbing algorithm for optimizing Convolutional Neural Networks (CNNs) on the CIFAR-100 dataset. The proposed model utilizes a population of c…
Modeling Managerial Search Behavior based on Simon's Concept of Satisficing
Computational models of managerial search often build on backward-looking search based on hill-climbing algorithms. Regardless of its prevalence, there is some evidence that this family of algorithms does not universally…
Decision MakingDiversified Late Acceptance Search
The well-known Late Acceptance Hill Climbing (LAHC) search aims to overcome the main downside of traditional Hill Climbing (HC) search, which is often quickly trapped in a local optimum due to strictly accepting only non…
DiversityClustering by Hill-Climbing: Consistency Results
We consider several hill-climbing approaches to clustering as formulated by Fukunaga and Hostetler in the 1970's. We study both continuous-space and discrete-space (i.e., medoid) variants and establish their consistency.
Clustering