Filters








123,746 Hits in 4.5 sec

Probability type inference for flexible approximate programming

Brett Boston, Adrian Sampson, Dan Grossman, Luis Ceze
2015 SIGPLAN notices  
Solver-aided type inference lets the programmer specify the correctness probability on only some variables-program outputs, for example-and automatically fills in other types to meet these specifications  ...  In approximate computing, programs gain efficiency by allowing occasional errors. Controlling the probabilistic effects of this approximation remains a key challenge.  ...  We propose DECAF (DECAF, an Energy-aware Compiler to make Approximation Flexible), a type-based approach to controlling quality in approximate programs.  ... 
doi:10.1145/2858965.2814301 fatcat:rbotzddksjazljfz7jj6reikfu

Probability type inference for flexible approximate programming

Brett Boston, Adrian Sampson, Dan Grossman, Luis Ceze
2015 Proceedings of the 2015 ACM SIGPLAN International Conference on Object-Oriented Programming, Systems, Languages, and Applications - OOPSLA 2015  
Solver-aided type inference lets the programmer specify the correctness probability on only some variables-program outputs, for example-and automatically fills in other types to meet these specifications  ...  In approximate computing, programs gain efficiency by allowing occasional errors. Controlling the probabilistic effects of this approximation remains a key challenge.  ...  We propose DECAF (DECAF, an Energy-aware Compiler to make Approximation Flexible), a type-based approach to controlling quality in approximate programs.  ... 
doi:10.1145/2814270.2814301 dblp:conf/oopsla/BostonSGC15 fatcat:qmugfpw4b5dw3pe7dvhn6naj7a

Universal Marginaliser for Deep Amortised Inference for Probabilistic Programs [article]

Robert Walecki, Kostis Gourgoulias, Adam Baker, Chris Hart, Chris Lucas, Max Zwiessele, Albert Buchard, Maria Lomeli, Yura Perov, Saurabh Johri
2019 arXiv   pre-print
To this end, we present the Universal Marginaliser (UM), a novel method for amortised inference, in PPL.  ...  We show how combining samples drawn from the original probabilistic program prior with an appropriate augmentation method allows us to train one neural network to approximate any of the corresponding conditional  ...  The flexible design allows the method to automatically build the network architecture for different types of probabilistic programs.  ... 
arXiv:1910.07474v1 fatcat:s5okn32phvg3diyoisfwnuokbi

Automatic Sampler Discovery via Probabilistic Programming and Approximate Bayesian Computation [chapter]

Yura Perov, Frank Wood
2016 Lecture Notes in Computer Science  
Specifically, we learn the procedure code of samplers for one-dimensional distributions.  ...  We formulate a Bayesian approach to this problem by specifying an adaptor grammar prior over probabilistic program code, and use approximate Bayesian computation to learn a program whose execution generates  ...  program for learning conditional sampler program text for Bernoulli(θ) in Figure 2 shows an example of this kind of approximation.  ... 
doi:10.1007/978-3-319-41649-6_27 fatcat:3jbkfa2t7bhubph2h6t6wl4z6e

Learning Probabilistic Programs [article]

Yura N. Perov, Frank D. Wood
2014 arXiv   pre-print
We also introduce a new notion of probabilistic program compilation and show how the same machinery might be used in the future to compile probabilistic programs for efficient reusable predictive inference  ...  We develop a technique for generalising from data in which models are samplers represented as program text.  ...  Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation heron.  ... 
arXiv:1407.2646v1 fatcat:gzapxdw7xvbk7m2eodqjjjpgwm

Static Analysis for Probabilistic Programs [article]

Ryan Bernstein
2019 arXiv   pre-print
Probabilistic programming is a powerful abstraction for statistical machine learning.  ...  We provide technical background for static analysis and probabilistic programming, suggest a functional taxonomy for probabilistic programming languages, and analyze the applicability of major ideas in  ...  These methods could potentially be extended to handle this flexibility by making conservative approximations. ! !  ... 
arXiv:1909.05076v1 fatcat:4rld375ztvbidm2nbqjpkbqzwy

Flexible Probabilistic Modeling for Search Based Test Data Generation

Robert Feldt, Shin Yoo
2020 Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops  
In particular, Probabilistic Programming languages (PPLs) and Genetic Programming (GP) should be investigated since they allow for very flexible probabilistic modelling.  ...  In this short paper, we present how some existing SBST techniques can be viewed from this perspective and then propose additional techniques for flexible generative modelling the community should consider  ...  Thus the more choices of a specific type that had been made the more different the probability distribution could be.  ... 
doi:10.1145/3387940.3392215 dblp:conf/icse/FeldtY20 fatcat:4xdujt2gy5budmkqh4bykkcejm

Probabilistic machine learning and artificial intelligence

Zoubin Ghahramani
2015 Nature  
any computable probability distribution as a probabilistic program [48] .  ...  Gaussian processes (GPs) are a very flexible nonparametric model for unknown functions, and are widely used for regression, classification, and many other applications that require inference on functions  ...  Gaussian processes (GPs) are a very flexible nonparametric model for unknown functions, and are widely used for regression, classification, and many other applications that require inference on functions  ... 
doi:10.1038/nature14541 pmid:26017444 fatcat:sw42v3vzcraj3mhimxr4w2g6du

Inference Over Programs That Make Predictions [article]

Yura Perov
2018 arXiv   pre-print
This abstract extends on the previous work (arXiv:1407.2646, arXiv:1606.00075) on program induction using probabilistic programming.  ...  It describes possible further steps to extend that work, such that, ultimately, automatic probabilistic program synthesis can generalise over any reasonable set of inputs and outputs, in particular in  ...  To make the priors over programs more flexible, we suggest to use non-parametric [6, 11] , hierarchical [9] priors.  ... 
arXiv:1810.01190v1 fatcat:mur6mjffbza4xp2dbu6aattzbu

Applications of Probabilistic Programming (Master's thesis, 2015) [article]

Yura N Perov
2020 arXiv   pre-print
In Chapter 3, we describe a way to facilitate sequential Monte Carlo inference in probabilistic programming using data-driven proposals.  ...  the facilitation of sequential Monte Carlo inference with help of data-driven proposals.  ...  The choice of distribution type for q(x t |η) is flexible. As mentioned, it should have the same structure of random choices as the prior p(x t |ρ t ).  ... 
arXiv:1606.00075v2 fatcat:bwargqfv25fjplh53spm7aitbe

Functional programming for modular Bayesian inference

Adam Ścibior, Ohad Kammar, Zoubin Ghahramani
2018 Proceedings of the ACM on Programming Languages  
We present an architectural design of a library for Bayesian modelling and inference in modern functional programming languages.  ...  In this program, we observe that the lawn is wet, which can be caused by either rain or the sprinkler, and try to infer the probability that it was raining.  ...  We thank David Tolpin for providing Anglican and WebPPL scripts we used for benchmarking and Jeremy Yallop for his help with using the OCaml type system.  ... 
doi:10.1145/3236778 dblp:journals/pacmpl/ScibiorKG18 fatcat:g6tspxtoh5gfvho4iy6mdxu5ay

Propositional and Relational Bayesian Networks Associated with Imprecise and Qualitative Probabilistic Assesments [article]

Fabio Gagliardi Cozman, Cassio Polpo de Campos, Jaime Ide, Jose Carlos Ferreira da Rocha
2012 arXiv   pre-print
New exact and approximate inference algorithms based on multilinear programming and iterated/loopy propagation of interval probabilities are presented; their superior performance, compared to existing  ...  This paper investigates a representation language with flexibility inspired by probabilistic logic and compactness inspired by relational Bayesian networks.  ...  The price to pay for all this flexibility is computational complexity and, often, inferential vacuousness -inferences typically lead to probability intervals, and often these intervals are quite wide.  ... 
arXiv:1207.4121v1 fatcat:jw2kwnvjirbrpduat6tmstmxm4

Probabilistic (logic) programming concepts

Luc De Raedt, Angelika Kimmig
2015 Machine Learning  
A multitude of different probabilistic programming languages exists today, all extending a traditional programming language with primitives to support modeling of complex, structured probability distributions  ...  While doing so, we focus on probabilistic extensions of logic programming languages such as Prolog, which have been considered for over 20 years.  ...  Acknowledgements The authors are indebted to Bernd Gutmann and Ingo Thon for participating in many discussions, and contributing several ideas during the early stages of the research that finally led to  ... 
doi:10.1007/s10994-015-5494-z fatcat:6bgcvas4lfed5invj2jjb7wfmy

Page 914 of Psychological Abstracts Vol. 83, Issue 3 [page]

1996 Psychological Abstracts  
83: 7073-7079 probability orderings or very specific assumptions on the nature of atoms in the probability space. To address these concerns, an approximate representability framework is developed.  ...  Two computer programs are included with “Analyzing Inter- action.” The SDIS program checks and prepares sequential data for efficient analysis.  ... 

Probabilistic Programming Concepts [article]

Luc De Raedt, Angelika Kimmig
2013 arXiv   pre-print
A multitude of different probabilistic programming languages exists today, all extending a traditional programming language with primitives to support modeling of complex, structured probability distributions  ...  Each of these languages employs its own probabilistic primitives, and comes with a particular syntax, semantics and inference procedure.  ...  Acknowledgements The authors are indebted to Bernd Gutmann and Ingo Thon for participating in many discussions, and contributing several ideas during the early stages of the research that finally led to  ... 
arXiv:1312.4328v1 fatcat:2zmu2qve2rbapnvsm7zesbbzwy
« Previous Showing results 1 — 15 out of 123,746 results