Confirmed Talks

• Keynote Address—Building the Ultimate Enterprise Computation Platform

  Creating an effective enterprise computation platform capable of empowering individual researchers while at the same time driving organisation-wide efficiency requires many components to come together at once. This talk will outline some of Wolfram Research's vision of what those components are, and will explain how some of the recent and future enhancements to the Wolfram Language support this goal. The talk will showcase real-world solutions that apply these ideas.

• Software Development Projects with the Wolfram Cloud

  Integration with other software and services has been paramount to the evolution of software development. While this has been previously possible with Mathematica and the Wolfram Language, the Wolfram Cloud offers the most convenient platform for the integration of Wolfram Language code to date. You will be shown a variety of practical examples of this using our unique functions, specifically automatically generated web APIs, web-based forms, and scheduled tasks.

  Create your own automated tweeter, a questionnaire that automatically generates a daily report, or even run your website entirely from the Wolfram Cloud.

• Computation Meets Knowledge: How Can Interactivity Improve Knowledge Transfer?

  When learning and communicating, exploration is much more engaging than passive reading. This session will explore ways of using Wolfram's solution, the Computable Document Format (CDF), to develop communicative tools and documents. Learn how to:

  • Conceive enriched dashboards
  • Maximise efficiency of your applications
  • Universally deploy any Wolfram Language code
• The Internet of Things: Unifying Global Data Sources

  As the Internet of Things begins to document, connect, and control our world, the way we interact with it needs to be as flexible as the devices it connects. Through simple, practical examples, this session will explore issues with portability and interconnectivity of the Wolfram Language–based approaches. Examples will include basic LED displays, a magic mirror, and a Reddit reader. Some of the tools used will be the Wolfram Data Drop, Wolfram|Alpha, and the Wolfram Language on a Raspberry Pi.

• Insight and Prediction—Practical Machine Learning

  Machine learning is moving rapidly from cutting-edge research to being an everyday tool for data analysis. Having been used in such disparate disciplines as healthcare, fraud detection, and animal conservation, the algorithms that underpin the multitude of approaches to analysis must be rigorous and reliable. This session will explore how the use of automation affects the reliability of predictions. The suite of tools offered by the Wolfram Language for the instantaneous creation and use of machine learning classifiers will be demonstrated on a range of practical examples.

• Unexpected Uses of Network Analysis

  While graph theory is central to the analysis of networks, it finds uses in some unexpected places, from text analysis to engineering reliability. This talk will show simple example problems from a range of fields and show how simple applications of graph theory tools can yield useful insights.

• Brain-Inspired Algorithms in Mathematica for Retinal Image Analysis (Eindhoven only)

  Damage to the retina due to diabetes is a main cause of blindness. Early detection is key, especially with the worldwide diabetes epidemic, in the form of large-scale screenings. To automatically detect micro-bleeds and irregularities in retinal vessels and retinal landmarks, we have designed a suite of algorithms inspired by recent findings of functional mechanisms in the visual cortex. Multiscale, and in particular multi-orientation, techniques turn out to be extremely robust. This is aided by high-dimensional computing, with tools of differential geometry, linear algebra, and Lie group analysis, for which Mathematica turns out to be a superb interactive design language.

• Wolfram Solutions for Financial Markets (London only)

  This talk will demonstrate the use of Wolfram technology/Mathematica in the financial industry, and will discuss what makes the platform an attractive proposition for research analysts, risk managers, quantitative developers, or other finance professionals who constantly seek quick and accurate solutions to their pertinent computational needs. Both traditional financial topics and more recent data science applications will be presented.

• Risk Management with Mathematica and UnRisk (Zurich only)

  In the last few years, risk management has become more and more important for many financial institutions. Increasing regulatory pressure as well as the need for improved internal risk control have shown that smaller institutions also need sound solutions. The combination of Mathematica and UnRisk allows for risk management that works across multiple asset classes and a vast variety of risk factors to improve the understanding of the sources of risk.

• Solving Real-World Business Problems in the Classroom (Zurich only)

  • How to use the Wolfram Language in a classroom environment to solve real business cases
  • How to teach coding to business administration students with no CS background
  • Showing results (interactive tools etc.) from the projects of the seminar Applied Business Modelling and Analytics 2016
• Innovative Accounting Visualizations by Wolfram Mathematica for Law Firms, Auditors, and Tax Advisors: Invep.Quantum and Invep.CDR (Berlin only, in German)

  The unique interactive 3D visualisation of Mathematica, docked to the well-known industry solution INVEP, facilitates practitioners in accounting, financial auditing, and revocatory action analysis to follow and investigate conjectures regarding delayed filing of insolvency, defeat of creditors, illiquidity, suspicious transactions, and more at a glance. The creation of a forecast is also simplified in such a way. Thus even with the biggest accounting systems a clear view can be generated.

• The Power of Mathematica Visualizations: Intuitively Understanding Technical Correlations (Berlin only)

  Interpreting and understanding technical correlations are essential in engineering work. Virtual modelling and simulation environments like Wolfram SystemModeler allow the user to investigate multi-physical systems already in the early stage of a development phase. In order to improve such systems, the understanding of technical correlations is a crucial step. However, the standard representation of simulation data (diagrams, plots, etc.) is often confusing to the user because of its inhibition of deductive reasoning, and hence is not beneficial for intuitively understanding the system's behaviour.

  This session will show how to overcome this barrier by using the power of Mathematica visualisations. Therefore, this contribution presents three differently realized go-cart models in Wolfram SystemModeler as demonstrative technical systems. A Modelica model of a fuel-driven go-cart will be simulated and compared against two differently implemented electrified go-cart models, one without gear set and one with a two-speed variable transmission. The differences between the models will be analysed. The presentation starts with the abundance of simulation data in order to show exemplarily how Mathematica can help to obtain impressive insights on the eye-minded path when simulation data is visualised appropriately.

• The Evidence of Bayes' Theorem (Eindhoven only)

  Bayes' theorem is trivially proved in probability theory; however, the application of Bayes' theorem is often far from trivial. Objections about "subjective prior probabilities" are often voiced as an argument against its use. These opponents may be unaware that Pierre-Simon Laplace used Bayes' theorem to determine the mass of Saturn to within 1% of NASA's current best estimate—almost two hundred years ago! Apparently, a subjective prior is not synonymous with an arbitrary prior.

  The normalisation constant in Bayes' theorem is called the evidence. The evidence is even more obscure than the prior. It is the single most important number in Bayesian analysis. It validates, objectively, the quality of different models, while each model itself is a good fit to the data. Computing the evidence invokes numerical integrations, often in many dimensions. This is the arena where the Wolfram Language can demonstrate its true power.

  The evidence is the unique selling point of Bayes' theorem.

• Solving Image Processing Problems with Mathematica (Warsaw only)

  State-of-the-art 2D and 3D image processing functions in Mathematica allow one to solve many practical problems. In this talk there will be examples from computer vision, microscopy, computational photography, and machine learning.

Confirmed Speakers