Filters








14 Hits in 4.7 sec

OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning [article]

Anurag Ajay, Aviral Kumar, Pulkit Agrawal, Sergey Levine, Ofir Nachum
2021 arXiv   pre-print
primitive behaviors before using these primitives for downstream task learning.  ...  In addition to benefiting offline policy optimization, we show that performing offline primitive learning in this way can also be leveraged for improving few-shot imitation learning as well as exploration  ...  ACKNOWLEDGEMENTS We would like to thank Ben Eysenbach and Kamyar Ghasemipour for valuable discussions at different points over the course of this work.  ... 
arXiv:2010.13611v3 fatcat:7feoc4kyvfcehk4dgbh5jdnzz4

TRAIL: Near-Optimal Imitation Learning with Suboptimal Data [article]

Mengjiao Yang, Sergey Levine, Ofir Nachum
2021 arXiv   pre-print
the transition model to reparametrize the action space for sample-efficient imitation learning.  ...  We ask the question, is it possible to utilize such suboptimal offline datasets to facilitate provably improved downstream imitation learning?  ...  ACKNOWLEDGMENTS We thank Dale Schuurmans and Bo Dai for valuable discussions. We thank Justin Fu, Anurag Ajay, and Konrad Zolna for assistance in setting up evaluation tasks.  ... 
arXiv:2110.14770v1 fatcat:ifyrhmzd6bfdno2sastrot3lwi

Comparing Model-free and Model-based Algorithms for Offline Reinforcement Learning [article]

Phillip Swazinna, Steffen Udluft, Daniel Hein, Thomas Runkler
2022
Offline reinforcement learning (RL) Algorithms are often designed with environments such as MuJoCo in mind, in which the planning horizon is extremely long and no noise exists.  ...  We compare model-free, model-based, as well as hybrid offline RL approaches on various industrial benchmark (IB) datasets to test the algorithms in settings closer to real world problems, including complex  ...  In OPAL (Offline Primitive discovery for Accelerating offline reinforcement Learning) (Ajay et al., 2020) the authors address problems with long horizons and sparse rewards.  ... 
doi:10.48550/arxiv.2201.05433 fatcat:bhf5dnifn5bv7d2ou47e2puegy

The Large Hadron Collider

Alison Wright, Richard Webb
2007 Nature  
Ellis 302 The quest for the quark-gluon plasma P. Braun-Munzinger & J. Stachel PERSPECTIVE 310 The God particle et al. L. Lederman  ...  Lessons for the future What lessons can be learned from the LHC saga?  ...  With this table now seemingly replete, why is there still hunger for further discovery?  ... 
doi:10.1038/448269a fatcat:dejm23nyv5fuzlw2o7fgui74a4

Cyber-Physical Energy Systems Security: Threat Modeling, Risk Assessment, Resources, Metrics, and Case Studies

Ioannis Zografopoulos, Juan Ospina, Xiaorui Liu, Charalambos Konstantinou
2021 IEEE Access  
Thus, their vital importance, interoperability, and plurality of computing devices make them prominent targets for malicious attacks aiming to disrupt their operations.  ...  The overarching framework for modeling, simulating, assessing, and mitigating attacks in a CPS is illustrated using four representative attack scenarios targeting CPES.  ...  His research interests include the development of intelligent systems for electric power systems (EPS) and smart-grid applications, machine learning and reinforcement learning models for DER control, renewable  ... 
doi:10.1109/access.2021.3058403 fatcat:pbe7xz7sxfe6lg4tvifqsollgm

Cyber-Physical Energy Systems Security: Threat Modeling, Risk Assessment, Resources, Metrics, and Case Studies [article]

Ioannis Zografopoulos, Juan Ospina, XiaoRui Liu, Charalambos Konstantinou
2021 arXiv   pre-print
Thus, their vital importance makes them prominent targets for malicious attacks aiming to disrupt their operations.  ...  Cyber-physical systems (CPS) are interconnected architectures that employ analog, digital, and communication resources for their interaction with the physical environment.  ...  evaluation [54] Simulation-based impact evaluation [55] Deep reinforcement learning analysis [56] Stochastic modeling assessment [57] Quantitative attack impact evaluation on substations [58]  ... 
arXiv:2101.10198v2 fatcat:b3ty5fhly5bprmph2emxg77ibe

Sensors and Actuators in Smart Cities

Mohammad Hammoudeh, Mounir Arioua
2018 Journal of Sensor and Actuator Networks  
Author Contributions: Alex Adim Obinikpo and Burak Kantarci conceived and pursued the literature survey on deep learning techniques on big sensed data for smart health applications, reviewed the state  ...  The authors are grateful to the Middle East University, Amman, Jordan for the financial support granted to cover the publication fee of this research article. Author Contributions: Abdelhamied A.  ...  To this end, Data analysis has two aspects: offline machine learning for knowledge discovery and runtime compenents for applying the knowledge during the execution.  ... 
doi:10.3390/jsan7010008 fatcat:pt7nkf4oaraijkmsndohahqtnq

Workshop on Visualization in High-energy Physics : HEPVIS '96 [article]

Lucas Taylor, Carlo E Vandoni
1997
Although it is clear that we need both 2D and 3D, we still have to learn quite a bit on rendering 3D abstract objects, especially for very large detectors. The issue of product source, e.g."  ...  In fact, during the workshop we learned that the HEPVIS96 proceedings will be printed in plain black and white!.  ...  Acknowledgements We thank the participants and organisers of HEPVIS 96 for such a pleasant and informative conference, and the many colleagues with whom we discussed the frustrations of offline analysis  ... 
doi:10.5170/cern-1997-001 fatcat:gfleik7wlzcz5kxik2idd2ktze

Point-To-Point Communication

1994 The International Journal of Supercomputer Applications and High Performance Computing  
KTH Skolan för Elektro-och Systemteknik School of Electrical Engineering 2013 Denna bok innehåller alla rapporter från kandidatexjobbskursen på skolan för elektro-och systemteknik, KTH, som pågick under  ...  Av 50 studenter som framgångsrik avslutade kursen kommer 11 teknologer från Skolan för teknikvetenskap och två studenter från Skolan för informations-och kommunikationsteknik.  ...  ACKNOWLEDGMENT A big thanks to our instructor Per Zetterberg for his helpful tips, pointers and for providing the "FrameworkBachelorChild"-framework, which built the base for the application "Child Supervisor  ... 
doi:10.1177/109434209400800304 fatcat:fr4kdw72o5dl5fppqnlh5rixwq

20th CERN School of Computing [article]

Carlo E Vandoni
1997
Another advantage of HyperCard was that it was quite extensible through its relatively easy to learn scripting language HyperTalk.  ...  Credit for the first purpose-built hypertext system usually goes to Professor Andy van Dam at Brown University in Providence, USA for the mainframe based FRESS system [6].  ...  Many image formats are used to greater or lesser degrees in HEP, including: GIF, JPEG, TIFF, BMP, etc. for static images; MPEG, Quicktime, AVI for multiple images; and Postscript, CGM, IV, VRML for graphics  ... 
doi:10.5170/cern-1997-008 fatcat:6kssvocyvbcgzgnblwl7lmctoi

Liprotides: a New Class of Protein Lipid-Complexes

Jannik N. Pedersen, Henriette Kristina S. Frislev, Jan S. Pedersen, Daniel E. Otzen
2016 Biophysical Journal  
Recent examples of this development include an analyser crystal based alignment system for grazing incidence techniques, developing reflectivity capability, an offline sample stage for pre-experiment setup  ...  The set of coordinates is used for the calculation of the scattering intensity, either for an oriented particle (as shown in figure 1 ) for for randomly oriented systems.  ...  Understanding of the physical mechanism of this reinforcement effect is important for developing rubbery materials with high performance.  ... 
doi:10.1016/j.bpj.2015.11.3085 fatcat:eqeeb47atfc4lisnurb25n55gu

Web Usage Mining [chapter]

Yew-Kwong Woon, Wee-Keong Ng, Ee-Peng Lim
Web Mining  
ACKNOWLEDGMENTS The authors thank the two anonymous referees and Professor Anthony Scime for their valuable comments on the draft of this chapter.  ...  ACKNOWLEDGMENT We thank Excite, AlltheWeb.com and AltaVista for providing the Web query data sets without which we could not have conducted this research.  ...  Another important aspect of Web content mining is the usage of the Web as a data source for knowledge discovery.  ... 
doi:10.4018/978-1-59140-414-9.ch018 fatcat:ap6nd7exerg3xe2izd6porxocu

LANGUAGE IN INDIA Strength for Today and Bright Hope for Tomorrow

M Thirumalai, B Mallikarjun, Sam, B A Sharada, A R Fatihi, Lakhan, Marie Jennifer, S M Bayer, G Ravichandran, L Baskaran, Ramamoorthy, Swarna Assistant (+1 others)
unpublished
It also provides for a multicultural dialogic platform for different groups of people belonging to different cultures.  ...  I have analyzed My Parent's Bedroom by Uwem Akpan, based on Dijk's socio-cognitive models for Critical Discourse Analysis.  ...  Sundarsingh for his advice and guidance in my research work.  ... 
fatcat:csob3q2ctrcfbi3tk6x2tskmue

CYBER-PHYSICAL DISRUPTION MODELING, ANALYSIS, AND MANAGEMENT: AN EVOLUTIONARY OBJECT-PROCESS MODEL-BASED ROBUST SYSTEMS ENGINEERING APPROACH

Yaniv Mordecai
2016 unpublished
Liron Yedidsion from the Technion -Israel Institute of Technology of the refereeing committee, for their feedback on the methods and results of this research, and for their recognition of the merit of  ...  His broad knowledge and experience, his attitude to research and practice have made it possible for me to successfully achieve my goals.  ...  Learning can take place in design-time (offline) or during runtime (online).  ... 
doi:10.13140/rg.2.2.17049.16486 fatcat:vng3w24jjncnribr7tu6asqata