Many radiological studies can reveal the presence of several co-existing abnormalities, each one represented by a distinct visual pattern. In this article we address the problem of learning a distance metric for plain radiographs that captures a notion of "radiological similarity": two chest radiographs are considered to be similar if they share similar abnormalities. Deep convolutional neural networks (DCNs) are used to learn a low-dimensional embedding for the radiographs that is equipped with the desired metric. Two loss functions are proposed to deal with multi-labelled images and potentially noisy labels. We report on a large-scale study involving over 745,000 chest radiographs whose labels were automatically extracted from free-text radiological reports through a natural language processing system. Using 4,500 validated exams, we demonstrate that the methodology performs satisfactorily on clustering and image retrieval tasks. Remarkably, the learned metric separates normal exams from those having radiological abnormalities.

Managed multi-context systems (mMCSs) allow for the integration of heterogeneous knowledge sources in a modular and very general way. They were, however, mainly designed for static scenarios and are therefore not well-suited for dynamic environments in which continuous reasoning over such heterogeneous knowledge with constantly arriving streams of data is necessary. In this paper, we introduce reactive multi-context systems (rMCSs), a framework for reactive reasoning in the presence of heterogeneous knowledge sources and data streams. We show that rMCSs are indeed well-suited for this purpose by illustrating how several typical problems arising in the context of stream reasoning can be handled using them, by showing how inconsistencies possibly occurring in the integration of multiple knowledge sources can be handled, and by arguing that the potential non-determinism of rMCSs can be avoided if needed using an alternative, more skeptical well-founded semantics instead with beneficial computational properties. We also investigate the computational complexity of various reasoning problems related to rMCSs. Finally, we discuss related work, and show that rMCSs do not only generalize mMCSs to dynamic settings, but also capture/extend relevant approaches w.r.t.

Artificial intelligence holds great promise for everything from employee productivity to marketing campaign success. With early implementations of technologies such as predictive analytics, about half of the companies in three major business regions -- western Europe, Asia/Pacific and the United States -- plan to adopt some form of AI within the next five years. While there are significant obstacles that you and your business should consider, from stakeholder buy-ins to a lack of skilled workers, there are some simple tactics to help your enterprise gain a foothold with this technology and reap the benefits. Check out the IDC infographic below and be on your way to using AI to uncover new and unexpected insights.

There is less than just a handful 10,000 number of specialized talent in AI in the entire world. The war for AI talent, henceforth, would be ruthless enough to easily dwarf the challenge for spotting good software engineers. As it happens, you know that machine thinking has begun to usurp human thinking. Despite India's global dominance in offering cheap engineering talent at scale, young AI companies in India are looking for alternate ways to suck in what's available. Particularly for data scientists wherein the essence of AI skills lies - ranked 0.8 on Belong's talent supply index.

Political turmoil and hijinks abounded this week, but there were plenty of security antics playing out online, too. Researcher Sabri Haddouche released a suite of tricks and tools, collectively called Mailsploit, that allow you to send perfectly spoofed messages from more than a dozen popular email clients. And speaking of phishing, new research shows a spike in the use of HTTPS web encryption on phishing sites. Attackers want the green padlock that comes with HTTPS to make their phishing sites look more legitimate and persuasive to potential victims. At least the ad blocker Ghostery is working on using artificial intelligence to catch--and block--new types of ad-trackers more quickly.

Television programming in Spain has undergone a transformation over the past decade - changes driven partly by economics and partly by politics. Ever since the banking crisis of 2008, the country has been in a semi-constant state of political upheaval. A series of corruption scandals, inconclusive general elections and, more recently, Catalonia's run at independence have kept Spaniards glued to their televisions and pundits talking 24/7. That has given rise to a wave of political talk shows that the Spanish call tertulias. These programmes meet two important criteria, they provide political flashpoints that audiences seem to like and they're cheap to produce.

Jan Machacek is a passionate technologist who shares his expertise and passion for software as the editor of the Open Source Journal, regularly contributes to open source projects and speaks at conferences in the UK and abroad. Jan is the author of many open source projects (various Typesafe Activators, Reactive Monitor, Akka Patterns, Akka Extras, Scalad, Specs2 Spring, etc.), books and articles. It's all about Containers, Serverless and Reactive Programming right now! ProgSCon London will explore these trends through engaging talks delivered by leading industry experts. Several talks will also feature various aspect of Blockchain, Microservices and Big Data. If you are a software developer looking to sharpen your skills and learn from the best in the industry, then ProgSCon London 2017 is the place you need to be at!

Christina is audience development editor. After graduating from the University of Nottingham reading philosophy and theology in 2013, Christina joined a tech start-up specialising in mobile apps. She has a keen interest in the mobile platform and innovative tech.

International Business Machines' (IBM) research laboratory in Zurich, Switzerland has developed a new generic preprocessing building block that could make the speed by which machine learning algorithms can absorb new information faster. Such development is expected to largely benefit the booming AI industry. According to IBM Zurich mathematician Thomas Parnell, they have developed a generic solution to the AI learning process with a 10 times speedup. "To the best of our knowledge, we are first to have generic solution with a 10x speedup. Specifically, for traditional, linear machine learning models -- which are widely used for data sets that are too big for neural networks to train on -- we have implemented the techniques on the best reference schemes and demonstrated a minimum of a 10x speedup."

We derive a novel sensitivity analysis of input variables for predictive epistemic and aleatoric uncertainty. We use Bayesian neural networks with latent variables as a model class and illustrate the usefulness of our sensitivity analysis on real-world datasets. Our method increases the interpretability of complex black-box probabilistic models.