Distributional Reward Decomposition for Reinforcement Learning [PDF]
arXiv, 2019 Many reinforcement learning (RL) tasks have specific properties that can be leveraged to modify existing RL algorithms to adapt to those tasks and further improve performance, and a general class of such properties is the multiple reward channel. In those environments the full reward can be decomposed into sub-rewards obtained from different channels ...
arxiv
Developing reference plans for evaluating global clinical trials credentialing and PSQA systems
Journal of Applied Clinical Medical Physics, EarlyView.Abstract Purpose To develop a practical framework for creating a diverse set of validated reference plans (varying in complexity) and implement a workflow to introduce beam modeling, calibration, and delivery errors into the reference cohort to test and compare various dosimetry audit methodologies.
Fre'Etta M. D. Brooks +13 more
wiley +1 more source
Informativeness of Reward Functions in Reinforcement Learning [PDF]
arXivReward functions are central in specifying the task we want a reinforcement learning agent to perform. Given a task and desired optimal behavior, we study the problem of designing informative reward functions so that the designed rewards speed up the agent's convergence. In particular, we consider expert-driven reward design settings where an expert or
arxiv
"Every effort we put into the SAVINGS BOND PROGRAM will be richly rewarded . . ." [PDF]
, 1958 openalex +1 more source
Automated rapidplan model validation using Eclipse scripting API
Journal of Applied Clinical Medical Physics, EarlyView.Abstract RapidPlan offers efficiency gains and quality improvements in treatment planning. Prior to its use in the clinic, it requires an extensive validation procedure in which established clinical plans and those generated by the model are compared. The manual iterative nature of this process is resource intensive, as numerous iterations are required
Bradley Beeksma +2 more
wiley +1 more source
Lateral hypothalamic reward decreased by intragastric feeding: Self-determined “threshold” technique [PDF]
, 1967 Gary B. Mount, Bartley G. Hoebel
openalex +1 more source
The burden of burnout: Understanding its prevalence and organizational drivers in medical physics
Journal of Applied Clinical Medical Physics, EarlyView.Abstract Background Burnout is a work‐related syndrome characterized by increased levels of emotional exhaustion (EE) and depersonalization (DP) along with decreased levels of personal achievement. In the healthcare setting, higher burnout levels have been associated with negative impacts on personnel, an increased risk of errors, and a decrease in the
Deborah Schofield +4 more
wiley +1 more source
Persistence in punishment and extinction testing as a function of percentages of punishment and reward in training [PDF]
, 1967 Charles N. Uhl
openalex +1 more source
Journal of Applied Clinical Medical Physics, EarlyView.Abstract The well‐being of medical physicists can impact overall system performance, patient safety, and quality of patient care. There are limited formal assessments of factors contributing to physicists well‐being. Nine medical physicists at a US academic medical center were surveyed on 21 workplace factors, drawn from the National Academy of ...
Elizabeth Kwong +13 more
wiley +1 more source
Beyond Simple Sum of Delayed Rewards: Non-Markovian Reward Modeling for Reinforcement Learning [PDF]
arXivReinforcement Learning (RL) empowers agents to acquire various skills by learning from reward signals. Unfortunately, designing high-quality instance-level rewards often demands significant effort. An emerging alternative, RL with delayed reward, focuses on learning from rewards presented periodically, which can be obtained from human evaluators ...
arxiv