Category Ranking
98%
Total Visits
921
Avg Visit Duration
2 minutes
Citations
20
Article Abstract
Current unsupervised reinforcement learning methods often overlook reward nonstationarity during pre-training and the forgetting of exploratory behavior during fine-tuning. Our study introduces Self-Reference (SR), a novel add-on module designed to address both issues. SR stabilizes intrinsic rewards through historical referencing in pre-training, mitigating nonstationarity. During fine-tuning, it preserves exploratory behaviors, retaining valuable skills. Our approach significantly boosts the performance and sample efficiency of existing URL model-free methods on the Unsupervised Reinforcement Learning Benchmark, improving IQM by up to 17% and reducing the Optimality Gap by 31%. This highlights the general applicability and compatibility of our add-on module with existing methods.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1016/j.neunet.2025.107448 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Multi-Channel Temporal Interference Retinal Stimulation Based on Reinforcement Learning.
IEEE J Biomed Health Inform
September 2025
Retinal degenerative diseases such as age-related macular degeneration and retinitis pigmentosa cause severe vision impairment, while current electrical stimulation therapies are limited by poor spatial targeting precision. As a promising non-invasive alternative, the efficacy of temporal interference stimulation (TIS) for retinal targeting depends on optimized multi-electrode parameters. This study reconstructed a whole-head finite element model with detailed ocular structures and applied reinforcement learning (RL)-based multi-channel electrode parameter optimization to retinal stimulation.
View Article and Find Full Text PDFDetecting the Undetected: Machine Learning in Early Disease Diagnosis.
Basic Clin Pharmacol Toxicol
October 2025
Amity Institute of Biotechnology, Amity University, Noida, Uttar Pradesh, India.
Early detection of diseases is a critical pillar in advancing modern healthcare, offering timely interventions and better patient outcomes. This overview highlights a range of machine learning (ML) approaches that are transforming early disease diagnosis. We discuss how traditional supervised and unsupervised methods, alongside advanced deep learning and reinforcement learning techniques, are utilized to detect early disease markers, often before clinical symptoms appear.
View Article and Find Full Text PDFMultidimensional in silico evaluation of fluorine-18 radiopharmaceuticals: integrating pharmacokinetics, ADMET, and clustering for diagnostic stratification.
J Comput Aided Mol Des
September 2025
Department of Medical Physics and Biomedical Nanotechnologies, V.N. Karazin Kharkiv National University, 4 Svobody Sq., Kharkiv, 61022, Ukraine.
Fluorine-18-labeled radiopharmaceuticals are central to PET-based oncology imaging, yet comparative evaluations of their mechanistic behavior and diagnostic potential remain fragmented. In this study, we present a multidimensional in silico framework integrating pharmacokinetic modeling, structural ADMET prediction, and unsupervised clustering to systematically evaluate five widely used F-labeled PET radiopharmaceuticals: [F]FDG, [F]FET, [F]DOPA, [F]FMISO, and [F]FLT. Each radiopharmaceutical was simulated using a harmonized three-compartment model in COPASI to capture uptake dynamics under both normal and pathological conditions.
View Article and Find Full Text PDFProfiling mouse behavior with computational tools to assess age-dependent differences in associative learning.
Cell Rep Methods
August 2025
Cell-Type Mechanisms in Normal and Pathological Behavior Research Group, Neuroscience Research Program, Hospital del Mar Research Institute, 08003 Barcelona, Spain. Electronic address:
Second-order conditioning (SOC) enables animals to form associations between stimuli without direct reinforcement. In this study, we present a behavioral analysis pipeline that combines a light-tone SOC paradigm in mice with tools such as DeepLabCut, Keypoint-MoSeq, and DeepOF to evaluate responses across sex and age. Our results show that responses to the second-order stimulus (CS) specifically stem from its association with the first-order stimulus (CS).
View Article and Find Full Text PDFCase Report: Ischemic stroke in a young transgender woman due to unsupervised estrogen therapy.
Front Glob Womens Health
August 2025
Faculdade de Medicina, Universidade Federal do Mato Grosso do Sul (UFMS), Campo Grande, Mato Grosso do Sul, Brazil.
Introduction: Stroke is often associated with the elderly population, but recent epidemiological data indicate an increasing incidence among young adults. Among the risk factors, estrogenic hormone therapy (HT) has been linked to cerebrovascular events. This report presents the case of a transgender woman who suffered an ischemic stroke during the inappropriate use of HT, highlighting the importance of medical follow-up and risk assessment in gender-affirming therapy.
View Article and Find Full Text PDF