Filters








9 Hits in 6.6 sec

Reinforcement Learning in Practice: Opportunities and Challenges [article]

Yuxi Li
2022 arXiv   pre-print
Then we discuss challenges, in particular, 1) foundation, 2) representation, 3) reward, 4) exploration, 5) model, simulation, planning, and benchmarks, 6) off-policy/offline learning, 7) learning to learn  ...  In this article, we first give a brief introduction to reinforcement learning (RL), and its relationship with deep learning, machine learning and AI.  ...  DQN integrates Q-learning with DNNs, and utilizes the experience replay and a target network to stabilize the learning.  ... 
arXiv:2202.11296v2 fatcat:xdtsmme22rfpfn6rgfotcspnhy

The Sigma Cognitive Architecture and System: Towards Functionally Elegant Grand Unification

Paul S. Rosenbloom, Abram Demski, Volkan Ustun
2016 Journal of Artificial General Intelligence  
In this article, these four desiderata are motivated and explained, and then combined with the graphical architecture hypothesis to yield a rationale for the development of Sigma.  ...  Sigma as a whole is then analyzed in terms of how well the progress to date satisfies the desiderata.  ...  We would also like to thank our many collaborators on aspects of this effort, most of whom show up as authors or co-authors on one or more of the papers cited here.  ... 
doi:10.1515/jagi-2016-0001 fatcat:sja6zoxsyvb3fplose76shpoia

Autonomous and autonomic systems: with applications to NASA intelligent spacecraft operations and exploration systems

2010 ChoiceReviews  
We would like to thank the following people from NASA Goddard Space Flight Center who have contributed information on or some writings to material that covered flight software: Joe Hennessy, Elaine Shell  ...  We would like to thank them for their contributions. We have made liberal use of their work and their contributions to Agent-based research at NASA Goddard.  ...  The others will follow the new leader until they get to the object of interest and then swarm around and examine it.  ... 
doi:10.5860/choice.47-6277 fatcat:nlprjuayozhg7mp55fz2wkl4bi

Fly with me : algorithms and methods for influencing a flock [article]

Kathryn Long Genter
2017
Zavlanos et al. present a theoretical framework for controlling graph connectivity in mobile robot networks and consider flocking as an application of connectivity control [83] .  ...  For example, it might be possible to gain control of the flock behind an influencing agent if another influencing agent blocks the view of the updraft bird.  ...  In particular, it would be interesting to consider the following neighborhood models: • A weighted influence neighborhood model in which the orientation, speed, and position of farther agents is less accurate  ... 
doi:10.15781/t2445hv1g fatcat:rrlw6crz7zdtbboz44c7sihacq

Dagstuhl Reports, Volume 9, Issue 3, March 2019, Complete Issue [article]

2019
This implies that the user needs to trust the system to act according to the user's norms and values. Thus the system needs the capability of handling and reasoning about norms and values.  ...  We also need to understand what driver behaviors can be considered appropriate and safe in an automated vehicle context.  ...  The need to store and keep track of network information is different across different applications, what makes it harder to define a standard set of instructions.  ... 
doi:10.4230/dagrep.9.3 fatcat:2tpapmq2rfdsjpygzyvhw2l6ei

Transferable strategic meta-reasoning models

Michael Wunder
2013
To evaluate this model, I explore several experimental case studies that show how to use the framework to predict and respond to behavior using observed data, covering settings ranging from a small number  ...  For the first time, a confluence of advances in agent design, formation of massive online data sets of social behavior, and computational techniques have allowed for researchers to construct and learn  ...  ., 2010] to this game identifies a stable equilibrium and classifies agents as leaders or followers according to who initiates the equilibrium pattern.  ... 
doi:10.7282/t3jh3j8k fatcat:jrwgrekgknal3dh6gnkkdibsdi

Publishing Council: Intelligent Decision Support for Architecture and Integration of Next Generation Enterprises

Anton Železnikar, Ciril Baškovič, Cene Bavec, Matjan Krisper, Vladislav Rajkovič, Tatjana Welzer, Maggie Mcpherson, Pedro Isaias, Amjad Umar
unpublished
Architectures and integration of emerging next generation enterprises (NGEs) require a series of complex decisions.  ...  This paper describes an intelligent decision support environment that uses patterns, best practices, inferences, and collaboration for enterprise architecture and integration projects.  ...  We thank Pedro Isaias who motivated us to extend the original conference paper and to submit it for a special issue of the Informatica Journal.  ... 
fatcat:qkobpzh4xzayljby6c5yxuusua

A Novel Market-based Multi-agent System for Power Balance and Restoration in Power Networks [article]

(:Unkn) Unknown, University, My, Li Bai
2020
network can be dynamically constructed and reconfigured at run-time, which leads to a more nimble, flexible, and stable system.  ...  Based on case studies and simulation results, the proposed approach delivers more effective performance of contingencies response and better computation time efficiency as the scale of the power network  ...  A multi-leader multi-follower Stackelberg game is used to study the problem of energy trading network topology control for electric vehicles.  ... 
doi:10.34944/dspace/3447 fatcat:6lnr4gsckncqjpq2xn6nyzrzpa

Generic Reinforcement Learning Beyond Small MDPs [article]

Mayank Daswani, University, The Australian National, University, The Australian National
2016
The value- based cost allows for smaller representations, and its model-free nature allows for its extension to the function approximation setting, which has computational and representational advantages  ...  The primary motivation behind this thesis is to build FRL agents that work in practice, both for larger environments and larger classes of environments.  ...  If it attempts to move into a wall it receives -10. Eating a food pellet gains 10 and eating all the food on the map gains 100. Eating a ghost resets the ghost to the center of the map.  ... 
doi:10.25911/5d7637291a901 fatcat:m4qd7yrzczcohjub43bbmnrqya