A video released Friday showed former presidential candidate Hillary Clinton practicing how to duck a hug from her then opponent, President Donald Trump last year. Philippe Reines, a Democratic strategist who worked with Clinton's campaign during the 2016 election, posted the video on Twitter, which was filmed Sept. 24, 2016, two days before the first presidential debate. It showed Clinton practicing for her face-off with the then-candidate Trump, in which Reines opposite her, was portraying Trump. As a voice called out from the background, asking the two candidates to come on to the stage, both walked up, with Clinton extending her hand, but Reines (like Trump), opening his arms wide for a hug. Then laughter could be heard as Clinton tried to avoid the embrace.

The days of claiming artificial intelligence as a feature that set one startup or company apart from the others are over. These days, one would be hard-pressed to find any technology company attracting venture funding or partnerships that doesn't posit to use some form of machine learning. But for companies trying to innovate in healthcare using artificial intelligence, the stakes are considerably higher, meaning the hype surrounding the buzzword can be deflated far more quickly than in some other industry, where a mistaken algorithm doesn't mean the difference between life and death. Over the past five years, the number of digital health companies employing some form of artificial intelligence has dramatically increased. CB Insights tracked 100 AI-focused healthcare companies just this year, and noted 50 had raised their first equity rounds since January 2015.

Last year, in the height of the election season, the Obama administration quietly released a national strategic plan for artificial intelligence (AI) research and development. The plan was the beginning of a national effort to prepare Americans for a future with AI--a future some computer scientist believe our nation is ill-equipped to handle. AI has become a part of the American fabric for some time. Siri and Alexa are already taking orders, self-driving cars have hit some streets, and the concept of interconnectivity is now a reality through the Internet of Things. But experts assert that in order for the society to fully embrace AI, learning machines should not replace human workers, but complement them.

A new video shows an autonomous'robot pilot' successfully flying and landing a Boeing 737 in a simulator. The creepy robotic arm shifts around the cockpit as it rhythmically changes the air speed, adjusts the wing flaps and fires up the thrusters in preparation for landing. The project has been masterminded by the US Department of Defence's Defence Advanced Research Projects Agency (Darpa). The successful test takes the technology one step closer to transforming military planes and helicopters into autonomous flying machines. But the Alias robot goes a step further. For example, an array of cameras allows the robot to see all the cockpit instruments and read the gauges.

In this Dec. 13, 2016, file photo, Anthony Levandowski, head of Uber's self-driving program, speaks about their driverless car in San Francisco. SAN FRANCISCO -- Uber is threatening to fire a key executive accused of stealing self-driving car technology from a Google spin-off unless he waives his constitutional right against self-incrimination so the ride-hailing service can comply with a court order. The development raises the possibility that Uber may end up dumping Anthony Levandowski, whose expertise in robot-controlled cars is the main reason the ride-hailing company bought Levandowski's startup nine months ago. Until last month, Levandowski had been running Uber's self-driving car division. Although he no longer is doing that, he remains a vital part of Uber's effort to develop a fleet of robot cars so its service eventually will no longer have to rely on people to pick up passengers.

In March, the FAA noted that over 100,000 hobby drone owners had registered their machines since the year began, bringing the total in the US over 770,000. Owners have filed their non-commercial UAVs with the agency ever since the DoT passed a law in December 2015 that made registration mandatory. But a Washington, D.C. court has struck down that legislation, freeing just-for-fun drone owners from notifying the government of their purchases -- for good and ill. Model aircraft enthusiast John Taylor brought his case against the FAA back in January 2016, shortly after the regulations came in place. The DC court of appeals ruled (PDF) in his favor, effectively classifying non-commercial drones as model aircraft and subject to the FAA's 2012 Modernization and Reform Act, which prohibited the agency from making new laws restricting flying hobbyist craft. But the drone industry isn't celebrating this turn: Turns out, keeping track of owners and making sure they're trained to fly was useful for everyone.

An appeals court on Friday struck down a Federal Aviation Administration rule that required owners of drones used for recreation to register their craft. The ruling was a victory for hobbyists and a setback for the FAA, which cited safety concerns as it tried to tighten regulation of the fast-growing army of drone operators. Some pilots of commercial airliners have reported close calls with drones flying near airports. About 760,000 hobbyists have registered more than 1.6 million drones since 2015, and sales have skyrocketed. The FAA estimates that hobbyists will buy 2.3 million drones this year and 13 million by the end of 2020.

Abstract--Graph-based methods pervade the inference toolk-its of numerous disciplines including sociology, biology, neuroscience, physics, chemistry, and engineering. A challenging problem encountered in this context pertains to determining the attributes of a set of vertices given those of another subset at possibly different time instants. Leveraging spatiotemporal dynamics can drastically reduce the number of observed vertices, and hence the cost of sampling. Alleviating the limited flexibility of existing approaches, the present paper broadens the existing kernel-based graph function reconstruction framework to accommodate time-evolving functions over possibly time-evolving topologies. This approach inherits the versatility and generality of kernel-based methods, for which no knowledge on distributions or second-order statistics is required. Systematic guidelines are provided to construct two families of space-time kernels with complementary strengths. The first facilitates judicious control of regularization on a space-time frequency plane, whereas the second can afford time-varying topologies. Batch and online estimators are also put forth, and a novel kernel Kalman filter is developed to obtain these estimates at affordable computational cost. Numerical tests with real data sets corroborate the merits of the proposed methods relative to competing alternatives. A number of applications involving social, biological, brain, sensor, transportation, or communication networks call for efficient methods to infer the attributes of some vertices given the attributes of other vertices [1]. For example, in a social network with vertices and edges respectively representing persons and friendships, one may be interested in determining an individual's consumption trends based on those of their friends. This task emerges when sampling cost constraints, such as the impossibility to poll one country's entire population about political orientation, limit the number of vertices with known attributes. Existing approaches typically formulate this problem as the reconstruction of a function or signal on a graph [1]-[6], and rely on its smoothness with respect to the graph, in the sense that neighboring vertices have similar function values. This principle suggests, for instance, estimating one person's age by looking at their friends' age.

With the political world deeply focused on the question of whether the Trump Administration comprises a gang of Russian pawns, less attention has been devoted to more mundane questions such as: what ever happened to Trump's economic policy? As it happens, economists are keeping their eye on that ball, and their conclusion is that it's in a bad way. Most specifically, they recognize that Trump policy is aimed heavily at achieving annual economic growth of more than 3%. During the Presidential campaign, Trump promised growth of 3.5% a year, and sometimes even 4%. There's no disagreement that a sustained growth rate of this magnitude would be a significant achievement.

This year's PyData London conference was held in Bloomberg's offices on the 6th and 7th of May, with Tutorial Day on May 5th. As was the case with PyData Amsterdam 2017, I made the time to watch all of the talks from the conference, and write a blog post about the ones I found the most interesting. As I'm a huge fan of Random Forests, and consider them to pretty much be Data Science 101, I thoroughly enjoyed the talk given by Nathan Epstein from conference host Bloomberg. He gave a very good intuitive introduction to how the algorithm works, and also spoke about its advantages over Neural Networks - something very useful in a time when everyone is really gung-ho over Deep Learning and "AI". Ian Ozsvald, author of the great "High-Performance Python", together with Guzstav Belteki and Giles Weaver, presented a piece of research they did for the NHS, using data collected from ventilators used in neonatal wards.