Matches in SemOpenAlex for { <https://semopenalex.org/work/W2891790128> ?p ?o ?g. }
- W2891790128 abstract "Deep reinforcement learning (RL) has shown impressive results in a variety of domains, learning directly from high-dimensional sensory streams. However, when neural networks are trained in a fixed environment, such as a single level in a video game, they will usually overfit and fail to generalize to new levels. When RL models overfit, even slight modifications to the environment can result in poor agent performance. This paper explores how procedurally generated levels during training can increase generality. We show that for some games procedural level generation enables generalization to new levels within the same distribution. Additionally, it is possible to achieve better performance with less data by manipulating the difficulty of the levels in response to the performance of the agent. The generality of the learned behaviors is also evaluated on a set of human-designed levels. The results suggest that the ability to generalize to human-designed levels highly depends on the design of the level generators. We apply dimensionality reduction and clustering techniques to visualize the generators' distributions of levels and analyze to what degree they can produce levels similar to those designed by a human." @default.
- W2891790128 created "2018-09-27" @default.
- W2891790128 creator A5011623094 @default.
- W2891790128 creator A5020511097 @default.
- W2891790128 creator A5022307512 @default.
- W2891790128 creator A5025547740 @default.
- W2891790128 creator A5043102938 @default.
- W2891790128 creator A5077267552 @default.
- W2891790128 date "2018-06-28" @default.
- W2891790128 modified "2023-09-27" @default.
- W2891790128 title "Illuminating Generalization in Deep Reinforcement Learning through Procedural Level Generation" @default.
- W2891790128 cites W1572584819 @default.
- W2891790128 cites W1889013705 @default.
- W2891790128 cites W2028208918 @default.
- W2891790128 cites W2100097207 @default.
- W2891790128 cites W2116386744 @default.
- W2891790128 cites W2118916601 @default.
- W2891790128 cites W2296073425 @default.
- W2891790128 cites W2362143032 @default.
- W2891790128 cites W2490603777 @default.
- W2891790128 cites W2523246573 @default.
- W2891790128 cites W2530887700 @default.
- W2891790128 cites W2534314849 @default.
- W2891790128 cites W2565902248 @default.
- W2891790128 cites W2569196563 @default.
- W2891790128 cites W2603088459 @default.
- W2891790128 cites W2605801332 @default.
- W2891790128 cites W2624780181 @default.
- W2891790128 cites W2725489675 @default.
- W2891790128 cites W2727836376 @default.
- W2891790128 cites W2731828881 @default.
- W2891790128 cites W2733961795 @default.
- W2891790128 cites W2737215407 @default.
- W2891790128 cites W2749952662 @default.
- W2891790128 cites W2753316839 @default.
- W2891790128 cites W2766108544 @default.
- W2891790128 cites W2786036274 @default.
- W2891790128 cites W2789184669 @default.
- W2891790128 cites W2791696004 @default.
- W2891790128 cites W2797527950 @default.
- W2891790128 cites W2798854895 @default.
- W2891790128 cites W2806690340 @default.
- W2891790128 cites W2964043796 @default.
- W2891790128 hasPublicationYear "2018" @default.
- W2891790128 type Work @default.
- W2891790128 sameAs 2891790128 @default.
- W2891790128 citedByCount "80" @default.
- W2891790128 countsByYear W28917901282017 @default.
- W2891790128 countsByYear W28917901282018 @default.
- W2891790128 countsByYear W28917901282019 @default.
- W2891790128 countsByYear W28917901282020 @default.
- W2891790128 countsByYear W28917901282021 @default.
- W2891790128 countsByYear W28917901282022 @default.
- W2891790128 crossrefType "posted-content" @default.
- W2891790128 hasAuthorship W2891790128A5011623094 @default.
- W2891790128 hasAuthorship W2891790128A5020511097 @default.
- W2891790128 hasAuthorship W2891790128A5022307512 @default.
- W2891790128 hasAuthorship W2891790128A5025547740 @default.
- W2891790128 hasAuthorship W2891790128A5043102938 @default.
- W2891790128 hasAuthorship W2891790128A5077267552 @default.
- W2891790128 hasConcept C111030470 @default.
- W2891790128 hasConcept C119857082 @default.
- W2891790128 hasConcept C127413603 @default.
- W2891790128 hasConcept C134306372 @default.
- W2891790128 hasConcept C136197465 @default.
- W2891790128 hasConcept C154945302 @default.
- W2891790128 hasConcept C15744967 @default.
- W2891790128 hasConcept C177148314 @default.
- W2891790128 hasConcept C177264268 @default.
- W2891790128 hasConcept C177606310 @default.
- W2891790128 hasConcept C18903297 @default.
- W2891790128 hasConcept C199360897 @default.
- W2891790128 hasConcept C201995342 @default.
- W2891790128 hasConcept C22019652 @default.
- W2891790128 hasConcept C2780451532 @default.
- W2891790128 hasConcept C2780767217 @default.
- W2891790128 hasConcept C33923547 @default.
- W2891790128 hasConcept C41008148 @default.
- W2891790128 hasConcept C50644808 @default.
- W2891790128 hasConcept C542102704 @default.
- W2891790128 hasConcept C86803240 @default.
- W2891790128 hasConcept C97541855 @default.
- W2891790128 hasConceptScore W2891790128C111030470 @default.
- W2891790128 hasConceptScore W2891790128C119857082 @default.
- W2891790128 hasConceptScore W2891790128C127413603 @default.
- W2891790128 hasConceptScore W2891790128C134306372 @default.
- W2891790128 hasConceptScore W2891790128C136197465 @default.
- W2891790128 hasConceptScore W2891790128C154945302 @default.
- W2891790128 hasConceptScore W2891790128C15744967 @default.
- W2891790128 hasConceptScore W2891790128C177148314 @default.
- W2891790128 hasConceptScore W2891790128C177264268 @default.
- W2891790128 hasConceptScore W2891790128C177606310 @default.
- W2891790128 hasConceptScore W2891790128C18903297 @default.
- W2891790128 hasConceptScore W2891790128C199360897 @default.
- W2891790128 hasConceptScore W2891790128C201995342 @default.
- W2891790128 hasConceptScore W2891790128C22019652 @default.
- W2891790128 hasConceptScore W2891790128C2780451532 @default.
- W2891790128 hasConceptScore W2891790128C2780767217 @default.
- W2891790128 hasConceptScore W2891790128C33923547 @default.
- W2891790128 hasConceptScore W2891790128C41008148 @default.