A5027890741- Reinforcement Learning in Robotics
- Distributed Control Multi-Agent Systems
- Plant tissue culture and regeneration
- Privacy-Preserving Technologies in Data
- Caching and Content Delivery
- Plant nutrient uptake and metabolism
- Complex Network Analysis Techniques
- Modular Robots and Swarm Intelligence
- Opportunistic and Delay-Tolerant Networks
- Virology and Viral Diseases
- Topic Modeling
- Bacteriophages and microbial interactions
- Age of Information Optimization
- Neuroscience and Neuropharmacology Research
- Energy Load and Power Forecasting
- Plant Genetic and Mutation Studies
- Parasites and Host Interactions
- Advanced Radiotherapy Techniques
- Evacuation and Crowd Dynamics
- Web Data Mining and Analysis
- Fire Detection and Safety Systems
- Natural Language Processing Techniques
- DNA and Biological Computing
- Advanced Wireless Network Optimization
- Tree Root and Stability Studies
State Grid Corporation of China (China)
2024
Zhejiang University
2004-2023
Iowa State University
2011-2016
Southern Medical University
2003
Collagens require the hydroxylation of proline (Pro) residues in their triple-helical domain repeating sequence Xaa-Pro-Gly to function properly as a main structural component extracellular matrix animals at physiologically relevant conditions. The regioselective is catalyzed by specific prolyl 4-hydroxylase (P4H) posttranslational processing step.A recombinant human collagen type I α-1 (rCIα1) with high percentage hydroxylated prolines (Hyp) was produced transgenic maize seeds when...
With the rapid evolution of wireless mobile devices, there emerges an increased need to design effective collaboration mechanisms between intelligent agents gradually approach final collective objective by continuously learning from environment based on their individual observations. In this regard, independent reinforcement (IRL) is often deployed in multiagent alleviate problem a nonstationary environment. However, behavioral strategies IRL can be formulated only upon local observations...
The paper considers independent reinforcement learning (IRL) for multi-agent collaborative decision-making in the paradigm of federated (FL). However, FL generates excessive communication overheads between agents and a remote central server, especially when it involves large number or iterations. Besides, due to heterogeneity environments, multiple may undergo asynchronous Markov decision processes (MDPs), which will affect training samples model's convergence performance. On top...
The fifth-generation cellular networks (5G) have boosted the unprecedented convergence between information world and physical world. On other hand, empowered with enormous amount of data information, AI has been universally applied pervasive is believed to be an integral part 6G. Consequently, benefiting from advancement in communication technology AI, we boldly argue that conditions for collective intelligence (CI) will mature 6G era CI emerge among widely connected beings things....
Originating from entomology, stigmergy has provided an effective framework for swarm collaboration. Based on new discoveries astrocytes in regulating synaptic transmission the brain, this paper mapped mechanism into interaction mediated by propagation of calcium waves between synapses and investigated its characteristics advantages. Particularly, we have divided short-range that are not directly connected neurons three phases proposed a stigmergic learning model. In model, state change agent...
The paper considers independent reinforcement learning (IRL) for multi-agent decision-making process in the paradigm of federated (FL). We show that FL can clearly improve policy performance IRL terms training efficiency and stability. However, since parameters are trained locally aggregated iteratively through a central server FL, frequent information exchange incurs large amount communication overheads. To reach good balance between improving model's convergence reducing required...
In this letter, we propose a trustable policy collaboration scheme in the paradigm of multi-agent independent reinforcement learning (MAIRL). This is realized by directly mixing parameters homogeneous agents, for which an upper bound mixture metric derived to guarantee improvement. can update behavioral policies agents distributedly and further improve performance MAIRL. addition, develop practical implementation scheme, verify its effectiveness mixed-autonomy traffic control simulation...
To introduce a new method for generating the clinical target volume (CTV) from gross tumor (GTV) using geodesic distance calculation glioma.One glioblastoma patient was enrolled. The GTV and natural barriers were contoured on each slice of computer tomography (CT) simulation images. Then, graphic processing unit based parallel Euclidean transform used to generate CTV considering barriers. Three-dimensional (3D) visualization technique applied show delineation results. Speed operation...
In the paper, we studied solution of resource allocation problem with Coordinated Multi-Point(CoMP) in heterogeneous wireless networks. The utilization CoMP technology enhances channel quality users at edge cells, but it also brings more complexity to management network. To reduce this pressure, exploited a feedback-based method determine multi-layer Traditional Proportional Fairness(PF) achieves balance between system throughput and fairness. However, fairness among should be reconsidered...
The fifth-generation cellular networks (5G) has boosted the unprecedented convergence between information world and physical world. On other hand, empowered with enormous amount of data information, artificial intelligence (AI) been universally applied pervasive AI is believed to be an integral part six-generation (6G). Consequently, benefiting from advancement in communication technology AI, we boldly argue that conditions for collective (CI) will mature 6G era CI emerge among widely...
With the rapid evolution of wireless mobile devices, there emerges an increased need to design effective collaboration mechanisms between intelligent agents, so as gradually approach final collective objective through continuously learning from environment based on their individual observations. In this regard, independent reinforcement (IRL) is often deployed in multi-agent alleviate problem a non-stationary environment. However, behavioral strategies agents IRL can only be formulated upon...
The paper considers independent reinforcement learning (IRL) for multi-agent collaborative decision-making in the paradigm of federated (FL). However, FL generates excessive communication overheads between agents and a remote central server, especially when it involves large number or iterations. Besides, due to heterogeneity environments, multiple may undergo asynchronous Markov decision processes (MDPs), which will affect training samples model's convergence performance. On top...