Filters








61,986 Hits in 9.5 sec

Faster Statistical Model Checking by Means of Abstraction and Learning [chapter]

Ayoub Nouri, Balaji Raman, Marius Bozga, Axel Legay, Saddek Bensalem
2014 Lecture Notes in Computer Science  
This paper investigates the combined use of abstraction and probabilistic learning as a means to enhance statistical model checking performance.  ...  estimated by statistical model checking on the abstract model with respect to the concrete system.  ...  In Section 3, we present our contribution, that is the joint use of abstraction and learning as a means to speed-up statistical model checking.  ... 
doi:10.1007/978-3-319-11164-3_28 fatcat:ma4i5ji44vdmtbml3csmmkamdm

LALR: Theoretical and Experimental validation of Lipschitz Adaptive Learning Rate in Regression and Neural Networks [article]

Snehanshu Saha, Tejas Prashanth, Suraj Aralihalli, Sumedh Basarkod, T.S.B Sudarshan, Soma S Dhavala
2020 arXiv   pre-print
We propose a theoretical framework for an adaptive learning rate policy for the Mean Absolute Error loss function and Quantile loss function and evaluate its effectiveness for regression tasks.  ...  The framework is based on the theory of Lipschitz continuity, specifically utilizing the relationship between learning rate and Lipschitz constant of the loss function.  ...  ACKNOWLEDGEMENT The authors would like to thank the Science and Engineering Research Board (SERB)-DST, Government of of India for supporting this research (File SERB-EMR/ 2016/005687).  ... 
arXiv:2006.13307v1 fatcat:mml6g6yf7fh4xi22q2izhhj7ny

500+ Times Faster Than Deep Learning (A Case Study Exploring Faster Methods for Text Mining StackOverflow) [article]

Suvodeep Majumder, Nikhila Balaji, Katie Brey, Wei Fu, Tim Menzies
2018 arXiv   pre-print
Deep learning methods are useful for high-dimensional data and are becoming widely used in many areas of software engineering.  ...  This approach is over 500 times faster than deep learning (and over 900 times faster if we use all the cores on a standard laptop computer).  ...  Figure 1 : 1 Pseudo-code of Local Learning e model learns vector representation of a word (center word) by predicting surrounding words in a context window(c) by maximizing the mean of log probability  ... 
arXiv:1802.05319v1 fatcat:fooyzqezkfgbpcotoe2r6gqwtm

Reveal: A Formal Verification Tool for Verilog Designs [chapter]

Zaher S. Andraus, Mark H. Liffiton, Karem A. Sakallah
2008 Lecture Notes in Computer Science  
Concrete Design Abstract the Design Abstract Model Property holds Design is Design is Correct Correct Abstract Counterexample Bug Bug Trace Trace learn refine  ...  This paper examines the effect on Reveal's performance of the various available options for abstraction and refinement.  ...  If not, the counterexample is spurious, and is refuted in the abstract model by adding a blocking clause, similar to learning in SAT, and the process iterates.  ... 
doi:10.1007/978-3-540-89439-1_25 fatcat:33f3d7qdhbasdlyikpueucgpza

Page 23 of Psychological Abstracts Vol. 61, Issue 1 [page]

1979 Psychological Abstracts  
However, both copying and checking were faster with arrays of boxes than with arrays in which each alphanumeric character was separated by a small rising mark.  ...  Data appear to support the hypothesis of an inverted-U or inverted-J model relating performance (and learning) to muscular fatigue in motor control tasks. —Journal abstract. Visual Perception 192.  ... 

TF Boosted Trees: A scalable TensorFlow based framework for gradient boosting [article]

Natalia Ponomareva, Soroush Radpour, Gilbert Hendry, Salem Haykal, Thomas Colthurst, Petr Mitrichev, Alexander Grushetsky
2017 arXiv   pre-print
It is based on TensorFlow, and its distinguishing features include a novel architecture, automatic loss differentiation, layer-by-layer boosting that results in smaller ensembles and faster prediction,  ...  principled multi-class handling, and a number of regularization techniques to prevent overfitting.  ...  Once the Chief updates the model by adding a new layer, both gradients and quantile statistics become stale.  ... 
arXiv:1710.11555v1 fatcat:iuef46jawnh7zh5yuxpuaqwtby

Hazy

Arun Kumar, Feng Niu, Christopher Ré
2013 Communications of the ACM  
Acknowledgments The Hazy Research Group is a team of Ph.D., M.S., and undergraduate students, working under the supervision of Christopher Ré.  ...  Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the view of the above companies, DARPA, or the U.S. government  ...  Many state-of-the-art approaches to both of these challenges are largely statistical and combine rich databases with software driven by statistical analysis and machine learning.  ... 
doi:10.1145/2428556.2428570 fatcat:nvl7nqinpjbybg457uorkasu7e

Statistical Model Checking : An Overview [article]

Axel Legay, Benoit Delahaye
2010 arXiv   pre-print
The model checking problem for stochastic systems with respect to such logics is typically solved by a numerical approach that iteratively computes (or approximates) the exact measure of paths satisfying  ...  Another approach to solve the model checking problem is to simulate the system for finitely many runs, and use hypothesis testing to infer whether the samples provide a statistical evidence for the satisfaction  ...  Acknowledgments We would like to thanks our collaborators on the statistical model checking project: Sumit Jha, Marius Bozga, Saddeck Bensalem.  ... 
arXiv:1005.1327v1 fatcat:vwkiw5bjmfcddlkc43mlcxijai

Learning probabilistic models for model checking: an evolutionary approach and an empirical study

Jingyi Wang, Jun Sun, Qixia Yuan, Jun Pang
2018 International Journal on Software Tools for Technology Transfer (STTT)  
Our findings include that the effectiveness of learning may sometimes be limited and it is worth investigating how abstraction should be done properly in order to learn abstract models.  ...  ., model checking, model-based testing) rely on first obtaining a model of the system under analysis.  ...  Both learning and statistical model checking suffer from the rare event problem. The rare event problem is a major threat to the validity of both learning and statistical model checking.  ... 
doi:10.1007/s10009-018-0492-7 fatcat:4wy2t77vyvfotb5xmlf2wfrbsu

Page 1139 of Psychological Abstracts Vol. 42, Issue 8 [page]

1968 Psychological Abstracts  
—Many current stochastic models of learning imply stationarity of portions of individual response sequences.  ...  Letting n, = n,, the number of ranks in the “lower” X category, n, can be subtracted from each of the n, highest ranks, their mean can be found by the formula (n, + 1)/2, and n, readded to obtain Y’.  ... 

Improving active Mealy machine learning for protocol conformance testing

Fides Aarts, Harco Kuppens, Jan Tretmans, Frits Vaandrager, Sicco Verwer
2013 Machine Learning  
Using active learning, we learn a model M R of reference implementation R, which serves as input for a model based testing tool that checks conformance of implementation I to M R .  ...  We show how these tools can be used for learning models of and revealing errors in implementations, present the new notion of a conformance oracle, and demonstrate how conformance oracles can be used to  ...  Acknowledgements We thank Colin de la Higuera and Bernard Steffen for suggesting to use the learned reference model for answering equivalence queries, and Axel Belinfante for assisting us with JTorx.  ... 
doi:10.1007/s10994-013-5405-0 fatcat:ktmh6inwcfag7c7butm4ijnmbm

A General Framework for Mining Massive Data Streams

Pedro Domingos, Geoff Hulten
2003 Journal of Computational And Graphical Statistics  
Using this framework, we have successfully adapted several learning algorithms to massive data streams, including decision tree induction, Bayesian network learning, k-means clustering, and the EM algorithm  ...  In this extended abstract we identify some desiderata for such systems, and outline our framework for realizing them.  ...  Within this framework, we have designed and implemented massivestream versions of decision tree induction [1, 6] , Bayesian network learning [5] , k-means clustering [2] and the EM algorithm for mixtures  ... 
doi:10.1198/1061860032544 fatcat:zrimbutngvh47p3hr425fl4imq

Statistical Model Checking: An Overview [chapter]

Axel Legay, Benoît Delahaye, Saddek Bensalem
2010 Lecture Notes in Computer Science  
Another approach to solve the model checking problem is to simulate the system for finitely many executions, and use hypothesis testing to infer whether the samples provide a statistical evidence for the  ...  In this tutorial, we survey the statistical approach, and outline its main advantages in terms of efficiency, uniformity, and simplicity.  ...  Acknowledgments We would like to thanks our collaborators on the statistical model checking project: Sumit Jha, and Marius Bozga.  ... 
doi:10.1007/978-3-642-16612-9_11 fatcat:khfqnua7gfbstj7zq7d7vxlnsi

A General Framework for Mining Massive Data Streams [chapter]

Pedro Domingos, Geoff Hulten
2008 Knowledge Discovery from Sensor Data  
Using this framework, we have successfully adapted several learning algorithms to massive data streams, including decision tree induction, Bayesian network learning, k-means clustering, and the EM algorithm  ...  In this extended abstract we identify some desiderata for such systems, and outline our framework for realizing them.  ...  Within this framework, we have designed and implemented massivestream versions of decision tree induction [1, 6] , Bayesian network learning [5] , k-means clustering [2] and the EM algorithm for mixtures  ... 
doi:10.1201/9781420082333.ch2 fatcat:5yfldfwdqffkfldpokqkkzdena

Abstraction of Markov Population Dynamics via Generative Adversarial Nets [article]

Francesca Cairoli, Ginevra Carbone, Luca Bortolussi
2021 arXiv   pre-print
A strategy to reduce computational load is to abstract the population model, replacing it with a simpler stochastic model, faster to simulate.  ...  Compared to previous works, which rely on deep neural networks and Dirichlet processes, we explore the use of state of the art generative models, which are flexible enough to learn a full trajectory rather  ...  Acknowledgements This work has been partially supported by the Italian PRIN project "SEDUCE" n. 2017TWRCNB.  ... 
arXiv:2106.12981v1 fatcat:ozgfvnbk4ben5lk3ria54zq7uu
« Previous Showing results 1 — 15 out of 61,986 results