The proposed approach uses a map segmentation technique to decompose the environment map into smaller, tractable maps . A simple information gain function is computed to determine the best target region to search during eachiteration of the process . DDQN and A2C algorithms are extended with a stack ofLSTM layers and trained to generate optimal policies for the exploration andexploitation, respectively . We tested our approach in 3 different tasks against 4 baselines . The results demonstrate that our proposed approach is capable ofnavigating through randomly generated environments and covering more AoI in less time steps compared to the baselines. The results demonstrated that our proposal is able tonavigate through randomly generate environments and cover more AoIs inless time steps . The proposedapproach uses a new type of maps to explore and explore and exploit the AoIs. The proposed approaches are available to the highest levels of exploration and exploitation of these areas of interest. We hope to use this new types of maps in the next round of tests to improve our understanding of these types of operations. We have published a new version of this type of map.

Author(s) : Ashley Peake, Joe McCalmon, Yixin Zhang, Daniel Myers, Sarra Alqahtani, Paul Pauca

Links : PDF - Abstract

Code :
Coursera

Keywords : approach - map - environments - proposed - maps -

Leave a Reply

Your email address will not be published. Required fields are marked *