Some of the work in the thesis was previously presented in [Gal, 2015; Gal and Ghahramani, 2015a,b,c,d; Gal et al., 2016], but the thesis contains many new pieces of work as well. There are two factors at play when visualising uncertainty in dropout Bayesian neural networks: the dropout masks and the dropout probability of the first layer. Uncertainty depictions in my previous blog posts drew new dropout masks for each test point--which is equivalent to drawing a new prediction from the predictive distribution for each test point -2 \leq \x \leq 2 . More specifically, for each test point \x_i we drew a set of network parameters from the dropout approximate posterior \boh_{i} \sim q_\theta(\bo), and conditioned on these parameters we drew a prediction from the likelihood \y_i \sim p(\y \x_i, \boh_{i}) . Another important factor affecting visualisation is the dropout probability of the first layer.

Ben Dickson is a software engineer and the founder of TechTalks. Insider threats are the biggest cybersecurity threats to firms, organizations and government agencies. This is something you hear a lot at security conference keynotes and read about in data breach reports, white papers and surveys -- and these insider threats are becoming increasingly more difficult to detect and prevent, as well as more frequent. This seemingly unstoppable growth accentuates the problem and shortcomings of current solutions, and warrants the need for new defensive technologies to detect and stop the digital daggers aimed at our backs. Data science -- the application of mathematics, big data analytics and machine learning to extract knowledge and detect patterns -- is an emergent, advanced technology area that is proving its effectiveness in the realm of cybersecurity, including fighting insider threats.

A linear dynamical system (A,B,C,D) is equivalent to the system (TAT {-1}, TB, CT {-1}, D) for any invertible matrix T in terms of the behavior of the outputs. A little thought shows therefore that in its unrestricted parameterization the objective function cannot have a unique optimum. A common way of removing this redundancy is to impose a canonical form.

Technology is positioned to reshape the future of work. But without critical components--acceptance and adoption by people--it can never achieve its full impact. Why, when digital technology is so powerful, should organizations still prize the contributions of their people? Rick Lash is a senior client partner with the firm. While digital technology is enabling disruption--think Uber, Airbnb or Amazon--it's also facilitating an era of incredible opportunity.

With the advances of modern data storage technology, chips the size of your fingernail are capable of storing an entire library's worth of knowledge, so one thing you might think computers do better than people is remember things. But according to Google Inc.'s DeepMind team, the artificial intelligence research group that developed AlphaGo, that is not entirely true. In a new paper published in the journal Nature, DeepMind has outlined a process where it trained a neural network to have human-like memory, giving it not only the ability to store data, but also to recall that information and use it to solve novel problems. "Neural networks excel at pattern recognition and quick, reactive decision-making, but we are only just beginning to build neural networks that can think slowly – that is, deliberate or reason using knowledge," the DeepMind team wrote in a recent blog post. "For example, how could a neural network store memories for facts like the connections in a transport network and then logically reason about its pieces of knowledge to answer questions?" DeepMind calls its new method differentiable neural computers, and the team demonstrated its capabilities using the London Underground, one of the largest public transit systems in the world.

US Government report lays out guidance for AI use and regulation and puts regulating super AI's in the too hard bucket Artificial Intelligence (AI) research and development is starting to reach critical mass and new breakthroughs are being announced almost every day. Now a new report from the US Office of Science Technology Policy (OSTP), who advises Barak Obama directly on AI matters has prepared a new report on the technology which they see is increasingly poised to reshape the way we live and work. Titled Preparing for the Future of Artificial Intelligence the report makes 23 policy recommendations on a number of topics concerned with the best way to harness the power of machine learning and algorithm driven intelligence for the benefit of society. The OSTP position is that government has several roles to play in driving the direction of AI. Namely, "It should convene conversations about important issues and help to set the agenda for public debate. It should monitor the safety and fairness of applications as they develop, and adapt regulatory frameworks to encourage innovation while protecting the public. It should support basic research and the application of AI to public goods, as well as the development of a skilled, diverse workforce. And government should use AI itself, to serve the public faster, more effectively, and at lower cost."

When we talk about artificial intelligence (AI) most people still imagine robots who can talk, act, and behave (to a certain extent) like a human being -- like a C-3PO (Star Wars), sans the metallic look. Or maybe, a supercomputer that can read human behavior so well that it interacts seamlessly with us, while controlling the system -- like Hal 9000 (2001: A Space Odyssey) or Auto (Wall-E). While, arguably, we may not be there yet in terms of our command of AI, we are not that far. AI is definitely the direction tech development is taking, as evidenced by most recent trends, including the formation of a partnership by tech giants to push the frontier of AI. While we may not be nearing the Singularity, AI has taken leaps and bounds of improvement over the past few years alone.

Artificial intelligence (AI) is becoming more and more prevalent in almost every facet of people's day-to-day lives, from its ability to defeat world-class chess players to its implementation into self-driving cars. AI is hot, flashy, buzzworthy and is the future of many industries and applications, but many consumers don't realize the subtle way it is already influencing and shaping the world of retail. Thanks to an abundance of consumer data at their fingertips, retailers have slowly and subtly begun rolling out applications for many sectors of the retail industry, everything from using AI to offer better product recommendations, to chatbots that can carry on an (almost) lifelike conversation with consumers and help push them toward checkout, to being utilized in image recognition systems. "The pertinent question isn't necessarily when but where you'll see AI deployed. And the truth of the matter is that AI can benefit essentially every step and process of eCommerce, from site layout to personalization to -- and this part is extremely important -- customer happiness," Andy Narayanan, vice president of visual intelligence at Sentient, an artificial intelligence software provider, wrote for Total Retail in a piece back in July.

Facebook founder Mark Zuckerberg has received an offer from a famous face for his personal artificial-intelligence assistant: Robert Downey Jr. Mr. Zuckerberg turned to fans on Thursday for suggestions on who might provide the voice work for his AI and soon heard from Marvel's "Iron Man." The famous actor, whose character Tony Stark interacts with a computer named Jarvis, said he would be happy to oblige -- for charity. "I'll do in a heartbeat if Bettany gets paid and donates it to a cause of Cumberbatch's choosing…that's the right kind of STRANGE!" Mr. Downey Jr. said on the CEO's Facebook page.

DeepMind, an artificial intelligence firm that was acquired by Google in 2014 and is now under the Alphabet umbrella, has developed a computer than can refer to its own memory to learn facts and use that knowledge to answer questions. DeepMind says its new AI model, called a differentiable neural computer (DNC), can be fed with things like a family tree and a map of the London Underground network, and can answer complex questions about the relationships between items in those data structures. For example, you could get responses to questions like, "Starting at Bond street, and taking the Central line in a direction one stop, the Circle line in a direction for four stops, and the Jubilee line in a direction for two stops, at what stop do you wind up?" It's these networks that helped DeepMind's AlphaGo AI defeat world champions at the complex game of Go.