Hi,
apart from robotics and systems control requiring interaction what are the application fields of Reinforcement Learning,
What are the other fields where we can use reinforcement learning?
Wherever long-term decisions have to be optimized. Energy management, stock trading, ...
"systems control requiring interaction" basically cover all the use cases of RL, in one way or other even games, energy management, stock trading etc.
Some examples can be found at:
1. https://deepmind.com/blog/deepmind-ai-reduces-google-data-centre-cooling-bill-40/
2. https://arxiv.org/pdf/1803.11115.pdf
For Ad Networks: http://insticc.org/node/TechnicalProgram/icaart/presentationDetails/73955
Healthcare: http://insticc.org/node/TechnicalProgram/icaart/presentationDetails/75736
Thanks, guys gaming is included in the control, my concern is about interaction in RL !! can we imagine applying RL after adaptation to other fields is there any research or paper talking about this ??
This website contains an extensive list of RL applications to different domains, and the associated publications:
https://medium.com/@yuxili/rl-applications-73ef685c07eb
Using Reinforcement Learning to Design a Better Rocket Engine
https://blog.insightdatascience.com/using-reinforcement-learning-to-design-a-better-rocket-engine-4dfd1770497a
Robust control strategy for quadrotor helicopters
Article Deterministic Policy Gradient With Integral Compensator for ...
It is also part of system control, but I should say Reinforcement Learning can be used for the built environment (smart buildings & cities).
Article LightLearn: An adaptive and occupant centered controller for...
Chapter Reinforcement Learning for Intelligent Environments: A Tutorial
Article Reinforcement learning for demand response: A review of algo...
Hi @all,
I have had a deeper look into the basis of reinforcement learning over the last years. What is important is that there are actually two frameworks of reinforcement learning. The most widely used and adapted one is the discounted framework. However it is only applicable if the underlying Markov decision process (MDP) does not accumulate rewards too often, i.e. it works for games where at the end of the episode you get rewarded with +1/-1 but not for many practical applications which include profit (in terms of rewards) in every period. This is due to the average reward that gets accumulated and influences the state(-action) values with $\mathsf{Avg Reward} / (1-\gamma)$. Thus if it is $>0$ it very soon i) increases the state action values tremendously and ii) dilutes the estimated state-action values s.t. it is very hard to distinguish between good and bad actions, as the average reward is (in recurrent MDPs) the same for all states.
If you want reinforcement learning for such systems you have to go into average reward reinforcement learning. I am currently working on a paper for this, but you can look into the early works of Mahadevan, possibly starting with this one:
Mahadevan, S. (1996). Average reward reinforcement learning: Foundations, algorithms, and empirical results. Machine learning, 22(1-3), 159-195.
Actually, when talking about optimality in the sense of reinforcement learning as optimization problem, then the average reward reinforcement learning theoretically can find better policies than discounted reinforcement learning. However, from a practical point of view this still has to be shown, and that's what I am working on right now.
Interactive Lungs Auscultation with Reinforcement Learning Agent
Preprint Interactive Lungs Auscultation with Reinforcement Learning Agent
Multi-Agent Deep Reinforcement Learning for Multi-Object Tracker
Article Multi-Agent Deep Reinforcement Learning for Multi-Object Tracker
Optimal Longitudinal Control
Conference Paper LongiControl: A New Reinforcement Learning Environment
Traffic light control
Article Distributed traffic light control at uncoupled intersections...
Adaptive Gameplay
Chapter Improving Adaptive Gameplay in Serious Games Through Interac...
Self-Driving Bicycles
Article Toward Self-Driving Bicycles Using State-of-the-Art Deep Rei...
Autonomous Management of Energy-Harvesting IoT Nodes
Preprint Autonomous Management of Energy-Harvesting IoT Nodes Using D...
Online Energy Optimization of Flexible Manufacturing Systems Article An Artificial Intelligence Approach for Online Optimization ...
Emergency evacuation
Preprint Deep Q-Learning with Q-Matrix Transfer Learning for Novel Fi...
Dear Seied Mahdi Hashemi,
thank you very much for your support with the listing.
Would you like to add the related papers?
Multi-Agent Interaction
https://openai.com/blog/emergent-tool-use/
Preprint Emergent Tool Use From Multi-Agent Autocurricula
https://www.youtube.com/watch?v=kopoLzvh5jY&t=1s
Dear Roman Liessner
Yes , I do.
Article Supervised Fuzzy Reinforcement Learning for Robot Navigation
Article Neural inverse reinforcement learning in autonomous navigation
Article Mobile Robot Navigation with the Combination of Supervised L...
https://www.sciencedirect.com/science/article/abs/pii/S0019057807600329
Applications of reinforcement learning could be found in plasma physics, for example in calculations of tokamak disruptions.
Interactions can be interpreted as an impact on the actions in the case of controlled fuzzy Markov chains and fuzzy neural artificial networks.
Besides robotics ... games, multi-agent systems, recommender system, building systems control (ex: HVAC), or anything that you can roughly formalize as a MDP. I am applying to architectural design in my own research.
Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model
Preprint Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model
AI Learns to Park - Deep Reinforcement Learning
https://www.youtube.com/watch?v=VMp6pq6_QjI
In general, reinforcement learning is a framework one can apply to all stochastic, unknown, sequential decision making tasks. "Unknown", in this case, means the actual model, i.e. the Markov decision process, and hence the dynamics of the environment is not known.
In our research, we focus on Spatio-temporal applications of reinforcement learning. We claim that due to the availability of more and more data in smart cities, e.g. parking sensors, existing routing algorithms can be improved. However, even if real-time information about the environment is available, future development is often stochastic and unknown. Our vision paper provides information about this general idea:
Article Vision Paper: Reinforcement Learning in Smart Spatio-Tempora...
An example of a concrete application where present work is already applying reinforcement learning in Spatio-temporal tasks is the ride-sharing, or taxicab dispatching task.
Simultaneous Electric Powertrain Hardware and Energy Management Optimization of a Hybrid Electric Vehicle Using Deep Reinforcement Learning and Bayesian Optimization
Conference Paper Simultaneous Electric Powertrain Hardware and Energy Managem...
I think Rl can use in monitoring and in science we can use for new theory and we need new topic of ML soon.a new revolution in AI and ML.
-monitoring
-security
-new theorem in science like physics
I have finally brought together the paper showing where RL applications are applicable, or rather when discounted RL is not applicable, plus I established an algorithm that is applicable in such cases. Maybe this paper gives you more insight in the method:
Preprint Average Reward Adjusted Discounted Reinforcement Learning: N...
Volt-VAR Control in Power Distribution Systems with Deep Reinforcement Learning
Conference Paper Volt-VAR Control in Power Distribution Systems with Deep Rei...
Robust active flow control over a range of Reynolds numbers using an artificial neural network trained through deep reinforcement learning
Preprint Robust active flow control over a range of Reynolds numbers ...
This might be helpful:
https://medium.com/@yuxili/rl-applications-73ef685c07eb
Rebalancing the car-sharing system with reinforcement learning
Article Rebalancing the car-sharing system with reinforcement learning
Deep reinforcement learning for the control of microbial co-cultures in bioreactors
Article Deep reinforcement learning for the control of microbial co-...
IntelligentCrowd: Mobile Crowdsensing via Multi-Agent Reinforcement Learning
Article IntelligentCrowd: Mobile Crowdsensing via Multi-Agent Reinfo...
Robust flow control and optimal sensor placement using deep reinforcement learning
Preprint Robust flow control and optimal sensor placement using deep ...
Agent Coordination in Air Combat Simulation using Multi-Agent Deep Reinforcement Learning
Conference Paper Agent Coordination in Air Combat Simulation using Multi-Agen...
