Show simple item record

dc.contributor.authorSaldanha, Emilyen_US
dc.contributor.authorPraggastis, Brendaen_US
dc.contributor.authorBillow, Todden_US
dc.contributor.authorArendt, Dustin L.en_US
dc.contributor.editorJohansson, Jimmy and Sadlo, Filip and Marai, G. Elisabetaen_US
dc.date.accessioned2019-06-02T18:14:27Z
dc.date.available2019-06-02T18:14:27Z
dc.date.issued2019
dc.identifier.isbn978-3-03868-090-1
dc.identifier.urihttps://doi.org/10.2312/evs.20191168
dc.identifier.urihttps://diglib.eg.org:443/handle/10.2312/evs20191168
dc.description.abstractReinforcement learning (RL) is a branch of machine learning where an agent learns to maximize reward through trial and error. RL is challenging and data/compute intensive leading practitioners to become overwhelmed and make poor modeling decisions. Our contribution is a Visual Analytics tool designed to help data scientists maintain situation awareness during RL experimentation. Our tool allows users to understand which hyper-parameter values lead to better or worse outcomes, what behaviors are associated with high and low reward, and how behaviors evolve throughout training. We evaluated our tool through three uses cases using state of the art deep RL models demonstrating how our tool leads to RL situation awareness.en_US
dc.publisherThe Eurographics Associationen_US
dc.subjectHuman
dc.subjectcentered computing
dc.subjectVisualization systems and tools
dc.subjectComputing methodologies
dc.subjectReinforcement learning
dc.subjectComputational control theory
dc.titleReLVis: Visual Analytics for Situational Awareness During Reinforcement Learning Experimentationen_US
dc.description.seriesinformationEuroVis 2019 - Short Papers
dc.description.sectionheadersVolume, Simulation, and Data Reduction
dc.identifier.doi10.2312/evs.20191168
dc.identifier.pages43-47


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record