Videos from JuliaCon are now available online
2018
Previous editions: 2017 | 2016 | 2015 | 2014
Henrik Sykora

Budapest University of Technology and Economics


Poster

An effective method to investigate stochastic delayed systems using Julia

Modelling of stochastic effects in engineering is limited not only by the few existing mathematical methods but also by their low effectivity and high complexity. An easy-to-use Julia package is presented to investigate stochastic linear delayed systems including control and machine-tool vibrations.

Mathieu Besançon

GERAD, Montréal & INRIA Lille


Poster

Bi-level optimization for consumption predictability in smart grids

Increasing the proportion of renewable energy has become a priority for power grids, but their integration requires higher flexibility of consumption times. Using game theory and optimization, we design a framework allowing flexibility and privacy, while reducing costs for both consumer and supplier

Mauro Werder

ETH Zurich, Switzerland


Poster

BITE, a Baysian Ice Thickness Estimation model

Glaciers and ice caps are projected to dominate global sea level rise up to 2100. To estimate ice thickness maps of such glaciers using mostly surface data, we use a semi-physical forward model combined with a Bayesian inversion scheme. We leverage Julia to allow compute intense MCMC simulations.

Vivek Gadodia

Dravyaniti Consulting LLP


Poster

Built to Last- Developing Robust Trading and Investment Strategies

Life of a Trading strategy is unknown. 'Till how long will this work?'. As markets get more mature, with advent of ML and AI, the designer's job is getting tougher. Using rules to analyse big data by splitting it across Time-Frames, traders can gain significant edges, for all types of traders.

Laurent Heirendt, Ph.D.

University of Luxembourg


Poster

COBRA.jl - Gearing up for high-dimensional constraint-based modelling

Julia accelerates analyses of huge-scale biochemical networks with now more than a million biochemical reactions. Through the open-source high-performance COBRA.jl package, the analysis capabilities of the COnstraint-Based Reconstruction and Analysis community are lifted to another level.

Rob Blackwell

British Antarctic Survey


Poster

EchoJulia – a new approach to scientific echosounder data analysis

EchoJulia is a software library that is enabling new kinds of fisheries acoustic applications. From big data analysis of ship-based surveys to data collection from ocean gliders, we show how Julia is set to improve ecosystem estimates of Southern Ocean Antarctic Krill.

Pedro Ivo de Oliveira Filho

University College London


Poster

Enhancing Photovoltaic Solar Cell Manufacturing: Design and Scale-up of an Industrial AACVD under Uncertainty

Handling uncertainty in process design and optimisation is a crucial challenge in current research and in the industry. We believe that Julia has the potential to be used as a comprehensive modelling language to enable designers to develop new systems with greater confidence.

Dr. Zygmunt L. Szpak

Australian Institute for Machine Learning


Poster

Estimation of Nonlinear Errors-in-Variables Models for Computer Vision Applications

I will present an approximate maximum likelihood method that encapsulates a gamut of heteroscedastic errors-in-variables regression problems that often arise in computer vision applications. Julia's multiple dispatch and Unicode support facilitates a unique, concise and readable implementation.

R Gowers

University of Warwick


Poster

Extended Neurons with Noise in Julia

Most of the time, neurons in the cortex fire due to randomly arriving inputs. The irregular firing that results is not merely noise, but is thought to play an important role in cortical computation. I show how I have implemented stochastic simulations of this irregular firing in my research.

Andrew Collier

Exegetic Analytics


Poster

Getting Started with Bayesian Techniques in Julia

Bayesian techniques can seem challenging and esoteric. However, they promise a powerful alternative to frequentist statistics, with the possibility of extracting even more useful information from your data. This talk will help you get started with Bayesian analyses.

Petre Caraiani

Institute for Economic Forecasting, Romanian Academy


Poster

Introduction to Quantitative Macroeconomics using Julia

The title of the presentation comes from the my forthcoming book at one of the leading publishers, Academic Press, an imprint of Elsevier. The poster (based on the book) will introduce the audience to state-of-art computational techniques used in economics through examples and applications in Julia.

Ahan Sengupta

City of London School


Poster

Julia Robots on the Raspberry Pi

The talk is about remote control robots controlled by Julia on a Raspberry Pi. I will show how to do physical computing with Julia using the GPIO pins of the Pi.

Miguel Raz Guzmán Macedo

UNAM


Poster

Lessons for beginner developers: Arbitrary order generalized OrthogonalPolynomials.jl in 100 lines of Julia

Julia's capabilities are amazing for scientific computing - but how to begin designing your own libraries and packages? With OrthogonalPolynomials.jl, we will showcase the development of a production ready library in 100 lines of code - with a series of video walkthroughs on step by step design.

Renan Domingues

UFPR - Federal University of Paraná


Poster

LU update in the Simplex Method

We study the Simplex method, a linear optimization search method. The revised Simplex method was studied and implemented along with the LU update, aiming for the solution of large scale sparse problems. The method was implemented in the Julia Language and tested against available solvers.

beckerfloh@gmail.com

Karlsruhe Institute of Technology


Poster

PCE.jl - Towards a Julia Package for Polynomial Chaos Expansion

This poster announces PCE.jl, which is a Julia Package for uncertainty quantification (UQ) using a method called 'polynomial chaos expansion' (PCE). It offers an alternative approach to Monte-Carlo-based UQ, and is also more efficient for certain problem instances.

Will Tebbutt

University of Cambridge


Poster

Probabilistic Programming and Gaussian Processes in Julia

Deep Learning's popularity is partly derived from its composability. General probabilistic machine learning is, in fact, just as composable, but technically much more challenging. I present an interesting edge case where life is easy, and discuss how Julia is uniquely positioned to exploit this.

Chris Rackauckas

University of California, Irvine


Poster

The Hidden Information in Stochasticity: How Biological Organisms Use and Control Randomness

Stochasticity (randomness) in biochemical reactions exists. Things diffuse and randomly bump together to bind/unbind. Yet these chemicals form the basis of life. But biological organisms evolve to their surroundings. Have they evolved to use biochemical randomness? I will demonstrate how they do it.