Matches in SemOpenAlex for { <https://semopenalex.org/work/W600041284> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W600041284 endingPage "63" @default.
- W600041284 startingPage "57" @default.
- W600041284 abstract "<div class=abs_img><img src=[disp_template_path]/JRM/abst-image/00270001/07.jpg width=300 />Moderate-based reward generator</div> In conventional reinforcement learning, a reward function influences the learning results, and therefore, the reward function is very important. To design this function considering a task, knowledge of reinforcement learning is required. In addition to this, a reward function must be designed for each task. These requirements make the design of a reward function unfeasible. We focus on this problem and aim at realizing a method to generate a reward without the design of a special reward function. In this paper, we propose a universal evaluation for sensor inputs, which is independent of a task and is modeled on the basis of the indicator of pleasure and pain in biological organisms. This evaluation estimates the trend of sensor inputs based on the ease of input prediction. Instead of the design of a reward function, our approach assists a human being in learning how to interact with an agent and teaching it his/her demand. We recruited a research participant and attempted to solve the path planning problem. The results show that a participant can teach an agent his/her demand by interacting with the agent and the agent can generate an adaptive route by interacting with the participant and the environment. </span>" @default.
- W600041284 created "2016-06-24" @default.
- W600041284 creator A5067555059 @default.
- W600041284 creator A5088996104 @default.
- W600041284 date "2015-02-20" @default.
- W600041284 modified "2023-10-02" @default.
- W600041284 title "Self-Generation of Reward by Moderate-Based Index for Senor Inputsvspace" @default.
- W600041284 cites W1509319152 @default.
- W600041284 cites W1988708350 @default.
- W600041284 cites W2127412976 @default.
- W600041284 cites W2174817438 @default.
- W600041284 cites W2491525833 @default.
- W600041284 cites W2562182593 @default.
- W600041284 cites W2573524926 @default.
- W600041284 cites W2573714927 @default.
- W600041284 cites W2574439214 @default.
- W600041284 cites W2577129292 @default.
- W600041284 cites W2577208403 @default.
- W600041284 doi "https://doi.org/10.20965/jrm.2015.p0057" @default.
- W600041284 hasPublicationYear "2015" @default.
- W600041284 type Work @default.
- W600041284 sameAs 600041284 @default.
- W600041284 citedByCount "4" @default.
- W600041284 countsByYear W6000412842020 @default.
- W600041284 countsByYear W6000412842021 @default.
- W600041284 countsByYear W6000412842023 @default.
- W600041284 crossrefType "journal-article" @default.
- W600041284 hasAuthorship W600041284A5067555059 @default.
- W600041284 hasAuthorship W600041284A5088996104 @default.
- W600041284 hasBestOaLocation W6000412841 @default.
- W600041284 hasConcept C107457646 @default.
- W600041284 hasConcept C119857082 @default.
- W600041284 hasConcept C121332964 @default.
- W600041284 hasConcept C127413603 @default.
- W600041284 hasConcept C14036430 @default.
- W600041284 hasConcept C154945302 @default.
- W600041284 hasConcept C163258240 @default.
- W600041284 hasConcept C188116033 @default.
- W600041284 hasConcept C201995342 @default.
- W600041284 hasConcept C2780451532 @default.
- W600041284 hasConcept C2780992000 @default.
- W600041284 hasConcept C41008148 @default.
- W600041284 hasConcept C62520636 @default.
- W600041284 hasConcept C78458016 @default.
- W600041284 hasConcept C86803240 @default.
- W600041284 hasConcept C97541855 @default.
- W600041284 hasConceptScore W600041284C107457646 @default.
- W600041284 hasConceptScore W600041284C119857082 @default.
- W600041284 hasConceptScore W600041284C121332964 @default.
- W600041284 hasConceptScore W600041284C127413603 @default.
- W600041284 hasConceptScore W600041284C14036430 @default.
- W600041284 hasConceptScore W600041284C154945302 @default.
- W600041284 hasConceptScore W600041284C163258240 @default.
- W600041284 hasConceptScore W600041284C188116033 @default.
- W600041284 hasConceptScore W600041284C201995342 @default.
- W600041284 hasConceptScore W600041284C2780451532 @default.
- W600041284 hasConceptScore W600041284C2780992000 @default.
- W600041284 hasConceptScore W600041284C41008148 @default.
- W600041284 hasConceptScore W600041284C62520636 @default.
- W600041284 hasConceptScore W600041284C78458016 @default.
- W600041284 hasConceptScore W600041284C86803240 @default.
- W600041284 hasConceptScore W600041284C97541855 @default.
- W600041284 hasIssue "1" @default.
- W600041284 hasLocation W6000412841 @default.
- W600041284 hasOpenAccess W600041284 @default.
- W600041284 hasPrimaryLocation W6000412841 @default.
- W600041284 hasRelatedWork W2123899227 @default.
- W600041284 hasRelatedWork W2154793587 @default.
- W600041284 hasRelatedWork W260766989 @default.
- W600041284 hasRelatedWork W2959276766 @default.
- W600041284 hasRelatedWork W2961085424 @default.
- W600041284 hasRelatedWork W3074294383 @default.
- W600041284 hasRelatedWork W4206669594 @default.
- W600041284 hasRelatedWork W4295941380 @default.
- W600041284 hasRelatedWork W4319083788 @default.
- W600041284 hasRelatedWork W4378771262 @default.
- W600041284 hasVolume "27" @default.
- W600041284 isParatext "false" @default.
- W600041284 isRetracted "false" @default.
- W600041284 magId "600041284" @default.
- W600041284 workType "article" @default.
!["<div class=abs_img><img src=[disp_template_path]/JRM/abst-image/00270001/07.jpg width=300 />Moderate-based reward generator</div> In conventional reinforcement learning, a reward function influences the learning results, and therefore, the reward function is very important. To design this function considering a task, knowledge of reinforcement learning is required. In addition to this, a reward function must be designed for each task. These requirements make the design of a reward function unfeasible. We focus on this problem and aim at realizing a method to generate a reward without the design of a special reward function. In this paper, we propose a universal evaluation for sensor inputs, which is independent of a task and is modeled on the basis of the indicator of pleasure and pain in biological organisms. This evaluation estimates the trend of sensor inputs based on the ease of input prediction. Instead of the design of a reward function, our approach assists a human being in learning how to interact with an agent and teaching it his/her demand. We recruited a research participant and attempted to solve the path planning problem. The results show that a participant can teach an agent his/her demand by interacting with the agent and the agent can generate an adaptive route by interacting with the participant and the environment. </span>"](?object=%22%26lt%3Bdiv%20class%3Dabs_img%26gt%3B%26lt%3Bimg%20src%3D%5Bdisp_template_path%5D%2FJRM%2Fabst-image%2F00270001%2F07.jpg%20width%3D300%20%2F%26gt%3BModerate-based%20reward%20generator%26lt%3B%2Fdiv%26gt%3B%20In%20conventional%20reinforcement%20learning%2C%20a%20reward%20function%20influences%20the%20learning%20results%2C%20and%20therefore%2C%20the%20reward%20function%20is%20very%20important.%20To%20design%20this%20function%20considering%20a%20task%2C%20knowledge%20of%20reinforcement%20learning%20is%20required.%20In%20addition%20to%20this%2C%20a%20reward%20function%20must%20be%20designed%20for%20each%20task.%20These%20requirements%20make%20the%20design%20of%20a%20reward%20function%20unfeasible.%20We%20focus%20on%20this%20problem%20and%20aim%20at%20realizing%20a%20method%20to%20generate%20a%20reward%20without%20the%20design%20of%20a%20special%20reward%20function.%20In%20this%20paper%2C%20we%20propose%20a%20universal%20evaluation%20for%20sensor%20inputs%2C%20which%20is%20independent%20of%20a%20task%20and%20is%20modeled%20on%20the%20basis%20of%20the%20indicator%20of%20pleasure%20and%20pain%20in%20biological%20organisms.%20This%20evaluation%20estimates%20the%20trend%20of%20sensor%20inputs%20based%20on%20the%20ease%20of%20input%20prediction.%20Instead%20of%20the%20design%20of%20a%20reward%20function%2C%20our%20approach%20assists%20a%20human%20being%20in%20learning%20how%20to%20interact%20with%20an%20agent%20and%20teaching%20it%20his%2Fher%20demand.%20We%20recruited%20a%20research%20participant%20and%20attempted%20to%20solve%20the%20path%20planning%20problem.%20The%20results%20show%20that%20a%20participant%20can%20teach%20an%20agent%20his%2Fher%20demand%20by%20interacting%20with%20the%20agent%20and%20the%20agent%20can%20generate%20an%20adaptive%20route%20by%20interacting%20with%20the%20participant%20and%20the%20environment.%20%26lt%3B%2Fspan%26gt%3B%22){kind=link}