Imagine a future with self-navigating ships. As they ply the ocean autonomously their "digital captains" are far away on dry land, keeping watch remotely with mixed reality (MR) and artificial intelligence (AI) technologies. JRCS – a major Japanese maritime services company – believes it can make this a reality within the next 12 years. With the help of Microsoft, it has just launched an ambitious plan to digitally transform the global shipping industry. In a series of initial steps, JRCS is deploying MR, the Internet of Things (IoT), and AI to change how shipping crews are trained, how ships are maintained, and how navigational safety and standards are promoted and enforced.

China's support and focus for the emerging AI industry, as well as its goal of becoming a world leader in artificial intelligence by 2030, comes from a source that the country takes very seriously -- government policy. Thus, China's possible domination of the AI sphere within the next 12 years could be the result of the nation using its strengths and weaknesses as a means to propel itself into the forefront of intelligent tech. Kai-Fu Lee, an AI investor that helps tech startups get off the ground through his $1.8-billion dual-currency venture fund Sinovation Ventures, recently noted in a statement to WIRED that China's goal of becoming the world leader in AI-driven solutions by 2030 is achievable because the target is literally a policy from the state. According to Lee, China's government has the capability to implement policies that are followed to a fault by both the country's citizens and its business sector. The AI investor also noted that this system is difficult to replicate in other countries trying to dominate the AI field, such as the United States, due to the West's democracy-driven nature.

Deep learning, even if it is very successful nowadays, traditionally needs very large amounts of labeled data to perform excellent on the classification task. In an attempt to solve this problem, the one-shot learning paradigm, which makes use of just one labeled sample per class and prior knowledge, becomes increasingly important. In this paper, we propose a new one-shot learning method, dubbed MoVAE (Mixture of Variational AutoEncoders), to perform classification. Complementary to prior studies, MoVAE represents a shift of paradigm in comparison with the usual one-shot learning methods, as it does not use any prior knowledge. Instead, it starts from zero knowledge and one labeled sample per class. Afterward, by using unlabeled data and the generalization learning concept (in a way, more as humans do), it is capable to gradually improve by itself its performance. Even more, if there are no unlabeled data available MoVAE can still perform well in one-shot learning classification. We demonstrate empirically the efficiency of our proposed approach on three datasets, i.e. the handwritten digits (MNIST), fashion products (Fashion-MNIST), and handwritten characters (Omniglot), showing that MoVAE outperforms state-of-the-art one-shot learning algorithms.

We present an active learning architecture that allows a robot to actively learn which data collection strategy is most efficient for acquiring motor skills to achieve multiple outcomes, and generalise over its experience to achieve new outcomes. The robot explores its environment both via interactive learning and goal-babbling. It learns at the same time when, who and what to actively imitate from several available teachers, and learns when not to use social guidance but use active goal-oriented self-exploration. This is formalised in the framework of life-long strategic learning. The proposed architecture, called Socially Guided Intrinsic Motivation with Active Choice of Teacher and Strategy (SGIM-ACTS), relies on hierarchical active decisions of what and how to learn driven by empirical evaluation of learning progress for each learning strategy. We illustrate with an experiment where a simulated robot learns to control its arm for realising two kinds of different outcomes. It has to choose actively and hierarchically at each learning episode: 1) what to learn: which outcome is most interesting to select as a goal to focus on for goal-directed exploration; 2) how to learn: which data collection strategy to use among self-exploration, mimicry and emulation; 3) once he has decided when and what to imitate by choosing mimicry or emulation, then he has to choose who to imitate, from a set of different teachers. We show that SGIM-ACTS learns significantly more efficiently than using single learning strategies, and coherently selects the best strategy with respect to the chosen outcome, taking advantage of the available teachers (with different levels of skills).

Recommender systems are widely used in big information-based companies such as Google, Twitter, LinkedIn, and Netflix. A recommender system deals with the problem of information overload by filtering important information fragments according to users' preferences. In light of the increasing success of deep learning, recent studies have proved the benefits of using deep learning in various recommendation tasks. However, most proposed techniques only aim to target individuals, which cannot be efficiently applied in group recommendation. In this paper, we propose a deep learning architecture to solve the group recommendation problem. On the one hand, as different individual preferences in a group necessitate preference trade-offs in making group recommendations, it is essential that the recommendation model can discover substitutes among user behaviors. On the other hand, it has been observed that a user as an individual and as a group member behaves differently. To tackle such problems, we propose using an attention mechanism to capture the impact of each user in a group. Specifically, our model automatically learns the influence weight of each user in a group and recommends items to the group based on its members' weighted preferences. We conduct extensive experiments on four datasets. Our model significantly outperforms baseline methods and shows promising results in applying deep learning to the group recommendation problem.

JERUSALEM – Israel released details on Tuesday about what it described as an Iranian "air force" deployed in neighboring Syria, including civilian planes suspected of transferring arms, a signal that these could be attacked should tensions with Tehran escalate. Iran, along with Damascus and its big-power backer Russia, blamed Israel for an April 9 airstrike on a Syrian air base, T-4, that killed seven Iranian Revolutionary Guards Corps (IRGC) members. Iranian officials have promised unspecified reprisals. Israeli media ran satellite images and a map of five Syrian air bases allegedly used to field Iranian drones or cargo aircraft, as well as the names of three senior IRGC officers suspected of commanding related projects, such as missile units. The information came from the Israeli military, according to a wide range of television and radio stations and news websites.

The world of technology and finance is constantly growing and changing. With data on overload and limited processing power, the Fintech industry is finally ready to meet with AI and we're excited. Turns out we're not the only ones. In a recent Accenture poll, 79% of bankers believe that AI will revolutionize customer service and banking. The beginnings of these relationships are already starting to show their prevalence.

Molecular dynamics has long being seen as a computer simulation method used for studying physical movements of atoms and molecules interacting with each other and giving a view of dynamic evolution of the system. Deemed to be important for a routine study of macromolecules and their environments, molecular dynamics is now being combined with machine learning to get results in various nascent areas. The idea of combining molecular dynamics with ML dates back to 2008 when they were combined to improve the protein function recognition of a molecule. They treated molecules as dynamic entities and improved the ability of structure-based function prediction methods to specify possible functional sites. Since then it has been used for various functionalities including creation of hyper predictive computer models for drug discovery and simulation of infrared spectra, among others.

It has barely been 24 hours since the campaign began, and the smart and seamless, TicPods Free by Mobvoi have already raised USD 229,994 thanks to 2487 backers. Described as the most interactive earbuds by Mobvoi, the stylish Human-machine Interaction device offers an experience that includes ultimate touch controls, optimised Bluetooth connectivity and clear, crisp audio. Founded in 2012 by ex-Googlers, Mobvoi aspires to define the next generation of human-machine interaction. Their engineers strive to innovate tech products that seamlessly integrate cutting-edge technology into daily life. To date, Mobvoi has raised six rounds of funding, led by firms including SIG, Sequoia Capital, Zhenfund, Google and Volkswagen.

How plentiful are machine learning jobs in 2018? Machine Learning jobs are everywhere. USA has been the leader in machine learning, and in tech hubs like Silicon Valley, it seems as though every company has data scientists employed. The trend has spread to the rest of the country, and there are no indications that any of this is slowing down. Europe has been lagging behind in industry adoption, but is catching up in a big way.