JuliaCon 2020 | DynamicPPL: Stan-like Speed for Dynamic Probabilistic Models | Mohamed Tarek
We present DynamicPPL.jl, a modular library providing a lightning-fast infrastructure for probabilistic programming and Bayesian inference, used in Turing.jl. DynamicPPL enables Turing to have C/Stan-like speeds for Bayesian inference involving static and dynamic models alike. Beside run-time speed, DynamicPPL provides a user-friendly domain-specific language for defining and then querying probabilistic models.
We present the preliminary high-level design and features of DynamicPPL.jl (https://github.com/TuringLang/DynamicPPL.jl), a modular library providing a lightning-fast infrastructure for probabilistic programming, used as a backend for Turing.jl (https://github.com/TuringLang/Turing.jl). Beside a computational performance that is often close to or better than Stan, DynamicPPL provides an intuitive domain-specific language (DSL) that allows the rapid development of complex dynamic probabilistic programs. Being entirely written in Julia, a high-level dynamic programming language for numerical computing, DynamicPPL inherits a rich set of features available through the Julia ecosystem. Since DynamicPPL is a modular, stand-alone library, any probabilistic programming system written in Julia, such as Turing.jl, can use DynamicPPL to specify models and trace their model parameters. The main features of DynamicPPL are: 1) a meta-programming based DSL for specifying dynamic models using an intuitive tilde-based notation; 2) a tracing data-structure for tracking random variables in dynamic probabilistic models; 3) a rich contextual dispatch system allowing tailored behaviour during model execution; and 4) a user-friendly syntax for probabilistic queries. Finally, we show in a variety of experiments that DynamicPPL, in combination with Turing.jl, achieves computational performance that is often close to or better than Stan. Time Stamps:
00:00 Welcome!
00:10 Help us add time stamps or captions to this video! See the description for details.
Want to help add timestamps to our YouTube videos to help with discoverability? Find out more here: https://github.com/JuliaCommunity/YouTubeVideoTimestamps
Interested in improving the auto generated captions? Get involved here: https://github.com/JuliaCommunity/YouTubeVideoSubtitles
Видео JuliaCon 2020 | DynamicPPL: Stan-like Speed for Dynamic Probabilistic Models | Mohamed Tarek канала The Julia Programming Language
We present the preliminary high-level design and features of DynamicPPL.jl (https://github.com/TuringLang/DynamicPPL.jl), a modular library providing a lightning-fast infrastructure for probabilistic programming, used as a backend for Turing.jl (https://github.com/TuringLang/Turing.jl). Beside a computational performance that is often close to or better than Stan, DynamicPPL provides an intuitive domain-specific language (DSL) that allows the rapid development of complex dynamic probabilistic programs. Being entirely written in Julia, a high-level dynamic programming language for numerical computing, DynamicPPL inherits a rich set of features available through the Julia ecosystem. Since DynamicPPL is a modular, stand-alone library, any probabilistic programming system written in Julia, such as Turing.jl, can use DynamicPPL to specify models and trace their model parameters. The main features of DynamicPPL are: 1) a meta-programming based DSL for specifying dynamic models using an intuitive tilde-based notation; 2) a tracing data-structure for tracking random variables in dynamic probabilistic models; 3) a rich contextual dispatch system allowing tailored behaviour during model execution; and 4) a user-friendly syntax for probabilistic queries. Finally, we show in a variety of experiments that DynamicPPL, in combination with Turing.jl, achieves computational performance that is often close to or better than Stan. Time Stamps:
00:00 Welcome!
00:10 Help us add time stamps or captions to this video! See the description for details.
Want to help add timestamps to our YouTube videos to help with discoverability? Find out more here: https://github.com/JuliaCommunity/YouTubeVideoTimestamps
Interested in improving the auto generated captions? Get involved here: https://github.com/JuliaCommunity/YouTubeVideoSubtitles
Видео JuliaCon 2020 | DynamicPPL: Stan-like Speed for Dynamic Probabilistic Models | Mohamed Tarek канала The Julia Programming Language
Показать
Комментарии отсутствуют
Информация о видео
30 июля 2020 г. 23:30:01
00:25:11
Другие видео канала
![](https://i.ytimg.com/vi/VlykuUot6xM/default.jpg)
![JuliaCon 2020 | BinaryBuilder.jl - The Subtle Art of Binaries That Just Work](https://i.ytimg.com/vi/3IyXsBwqll8/default.jpg)
![StanCon 2020. Talk 12: Ben Goodrich. Bayesian Inference without Probability Density Functions](https://i.ytimg.com/vi/_wfZSvasLFk/default.jpg)
![Differentiable Programming with Julia by Mike Innes](https://i.ytimg.com/vi/LjWzgTPFu14/default.jpg)
![Elixir vs Scala - LUDWIK BUKOWSKI](https://i.ytimg.com/vi/k1ANjrqtOEI/default.jpg)
![JuliaCon 2020 | Doing Scientific Machine Learning (SciML) With Julia](https://i.ytimg.com/vi/QwVO0Xh2Hbg/default.jpg)
![Tutorial: Probabilistic programming - a categorical tutorial (Sam Staton)](https://i.ytimg.com/vi/JimCpEG0nts/default.jpg)
![The impact of differentiable programming: how ∂P is enabling new science in Julia](https://i.ytimg.com/vi/rF2QAJLM730/default.jpg)
![JuliaCon 2020 | Using VS Code for Julia development | David Anthoff](https://i.ytimg.com/vi/IdhnP00Y1Ks/default.jpg)
![JuliaCon 2020 | AlgebraicJulia: Applied Category Theory in Julia | James Fairbanks](https://i.ytimg.com/vi/7zr2qnud4XM/default.jpg)
![From Optimization to Probabilistic Programming](https://i.ytimg.com/vi/32rAwtTAGdU/default.jpg)
![JuliaCon 2019 | What's Bad About Julia | Jeff Bezanson](https://i.ytimg.com/vi/TPuJsgyu87U/default.jpg)
![Probabilistic Programming: The What, Why and How](https://i.ytimg.com/vi/cvD9DnTDxmY/default.jpg)
![JuliaCon 2020 | DynamicGrids.jl: high-performance spatial simulations in Julia | Rafael Schouten](https://i.ytimg.com/vi/MDvhIASXb6k/default.jpg)
![Julia: Is it better than Python? [Everything you need to know in 2020]](https://i.ytimg.com/vi/565EDmwCSZw/default.jpg)
![Julia and DifferentialEquations.jl : Chris Rackauckas, MIT](https://i.ytimg.com/vi/zJ3R6vOhibA/default.jpg)
![Working with images in Julia | Week 1 | 18.S191 MIT Fall 2020 | Grant Sanderson](https://i.ytimg.com/vi/DGojI9xcCfg/default.jpg)
![Andrew Gelman: Introduction to Bayesian Data Analysis and Stan with Andrew Gelman](https://i.ytimg.com/vi/T1gYvX5c2sM/default.jpg)
![Deep probabilistic Modelling with Pyro | Chi Nhan Nguyen](https://i.ytimg.com/vi/mtFBfOmdFQk/default.jpg)
![JuliaCon 2020 | Display, show and print -- how Julia's display system works | Fredrik Ekre](https://i.ytimg.com/vi/S1Fb5oNhhbc/default.jpg)