Reinforcement learning is concerned with identifying reward-maximizing behaviour policies in environments that are initially unknown. State-of-the-art reinforcement learning approaches, such as deep Q-networks, are model-free and learn to act effectively across a wide range of environments such as Atari games, but require huge amounts of data. Model-based techniques are more data-efficient, but need to acquire explicit knowledge about the environment. In this paper, we take a step towards using model-based techniques in environments with a high-dimensional visual state space by demonstrating that it is possible to learn system dynamics and the reward structure jointly. Our contribution is to extend a recently developed deep neural network for video frame prediction in Atari games to enable reward prediction as well. To this end, we phrase a joint optimization problem for minimizing both video frame and reward reconstruction loss, and adapt network parameters accordingly. Empirical evaluations on five Atari games demonstrate accurate cumulative reward prediction of up to 200 frames. We consider these results as opening up important directions for model-based reinforcement learning in complex, initially unknown environments.

Along with super-human strength and the ability to look great in chrome, robots can now add another talent to their box of tricks: self-healing. Roboticists have long aimed to use soft flexible materials, but these have a propensity to break making them unfit for purpose. A new technique can create soft robots that heal themselves when things go wrong. To prove the concept, researchers at the Free University of Brussels (VUB) in Belgium created a gripper, a robot hand and an artificial muscle, all with the ability to self-heal, out of rubbery polymers that look a bit like jelly. When ripped or cut they can knit back together completely.

Digital media entrepreneur Jessica Naziri recently made waves when she shared photos of her technology-themed wedding, complete with bouquets made of USB cables and portable charging packs for guests instead of sugar coated almonds. She wasn't the first bride to garner media attention for eccentric nuptials fueled by mobile apps and gadgets. Guests watched through headsets as a community manager from the San Francisco startup AltspaceVR officiated the ceremony. Time reported the couple spent $2,531 a piece on their headsets and computer. New technologies and social networks are completely revamping the wedding industry. Today, there are dozens of popular apps taking the place of wedding planners, while couples are registering for bitcoin or Airbnb bookings instead of fine china.

Dating apps have been slow to adopt video functionality. Big shot Tinder bought a video service in February but hasn't announced plans to add its functionality, while Hinge just included user-made movie clips for profiles -- a feature that Badoo launched last year. Today, the UK-based dating service is taking another step forward and adding video chat straight into the app, so users can move past text and talk in real-time. You'll have to exchange at least one message each before video chats become available, so at least users won't get spammed with calls from randos they haven't chatted with. Badoo is pitching it as an added layer of security, giving folks the opportunity to see if anybody sets off their creep alarms before meeting in person.

Because of complementary advances in natural language processing, machine learning, and image recognition, the range of tasks for which AI is well-suited is growing daily. And when a critical level of AI saturation is reached, we anticipate profound disruption in the world of work. Imagine a barrel perched on one end of a seesaw; it represents the capacity of AI. On the other side of the fulcrum sits a giant of a man, a stand-in for human labour. A garden hose runs into the barrel, slowly filling it with water.

For many years we've been helping clients with the application of Artificial Intelligence and Machine Learning. While it can appear complex and intimidating, it's not so bad once you understand a few key concepts. Artificial Intelligence (AI) is a sector by buzzwords and hype -- so how do you cut through the noise? Think of AI as the superset -- and everything else is a subset of it. Put another way, AI is the universe and things like Machine Learning, Neural Networks, and Deep Learning is the solar systems that it's made up of.

FOR robots to coexist amicably with us, they need to learn about personal space. A software upgrade could help droids navigate crowded places like malls without jostling people around them. Harmish Khambhaita and Rachid Alami at the University of Toulouse in France wanted to programme a robot to mimic human manners like stepping around one another, yielding to groups and respecting personal space. "The robot has to reason what the human would like to do and react," says Khambhaita. This may seem akin to what driverless cars do, but humans are tougher to predict than traffic.

The legendary Bigfoot is often described as the "missing link" between apes and man, but the Chilesaurus has an edge on Bigfoot: it is the missing link between herbivore dinosaurs and their carnivorous brethren. In a new study done by the University of Cambridge, Chilesaurus, which lived 150 million years ago, scientists now believe the dinosaur is an early member of the "Ornithischia," a "bird-hipped" group that includes dinosaurs such as the Stegosaurus and Iguanadon. Researchers found that the Chilesaurus has the same inverted hip structure of the Ornithischia group, which aids in complex digestive systems. But it also lacks the beak Ornithischia dinosaurs used for eating. "Chilesaurus almost looks like it was stitched together from different animals, which is why it baffled everybody," said Matthew Baron, a doctoral student in Cambridge University's Department of Earth Sciences and the paper's joint first author, in a statement.

CHICAGO (GenomeWeb) – To the world, Google may talk about artificial intelligence with the best of'em, but internally, the internet giant shies away from that term, particularly in life sciences and medicine. Nevertheless, the company continues to make progress in applying the technology to these markets, with DNA sequencing analysis being a particularly ideal application, Allen Day, a science advocate at Google, said at the recent Intelligent Systems for Molecular Biology European-Conference on Computational Biology (ISMB/ECCB) conference in Prague.

Every day we hear about Machine Learning and Big Data Analytics... 'United Parcel Service saves 39 million gallons of fuel after using Big Data Analytics to optimise fleet operations'; 'PayPal uses Machine Learning on Customer, Financial and Network data to combat fraud'; 'Amazon uses Machine Learning to discover'lowest price' for over 20 million products'... Machine learning, a subset of Artificial Intelligence (AI) is a method of data analysis that uses algorithms to iteratively learn from data and derive insights without being explicitly programmed. We can find examples of how Machine Learning is already a part of our daily lives -- like Google Maps, using location data from smartphones, analyses the speed of movement of traffic at any given time. Or like Amazon makes recommendations for products -- "customers who bought this item also bought". Behind all these lie complex algorithms that are continuously learning new data and refining outcomes. In banking for example, using client's financial data, risk preferences and desired target return, 'Robo-Advisers' provide personalised, algorithm driven portfolio management services without human supervision.