We construct a multilingual common semantic space based on distributional semantics, where words from multiple languages are projected into a shared space to enable knowledge and resource transfer across languages. Beyond word alignment, we introduce multiple cluster-level alignments and enforce the word clusters to be consistently distributed across multiple languages. We exploit three signals for clustering: (1) neighbor words in the monolingual word embedding space; (2) character-level information; and (3) linguistic properties (e.g., apposition, locative suffix) derived from linguistic structure knowledge bases available for thousands of languages. We introduce a new cluster-consistent correlational neural network to construct the common semantic space by aligning words as well as clusters. Intrinsic evaluation on monolingual and multilingual QVEC tasks shows our approach achieves significantly higher correlation with linguistic features than state-of-the-art multi-lingual embedding learning methods do. Using low-resource language name tagging as a case study for extrinsic evaluation, our approach achieves up to 24.5\% absolute F-score gain over the state of the art.

Partial label learning (PLL) aims to solve the problem where each training instance is associated with a set of candidate labels, one of which is the correct label. Most PLL algorithms try to disambiguate the candidate label set, by either simply treating each candidate label equally or iteratively identifying the true label. Nonetheless, existing algorithms usually treat all labels and instances equally, and the complexities of both labels and instances are not taken into consideration during the learning stage. Inspired by the successful application of self-paced learning strategy in machine learning field, we integrate the self-paced regime into the partial label learning framework and propose a novel Self-Paced Partial-Label Learning (SP-PLL) algorithm, which could control the learning process to alleviate the problem by ranking the priorities of the training examples together with their candidate labels during each learning iteration. Extensive experiments and comparisons with other baseline methods demonstrate the effectiveness and robustness of the proposed method.

In recent years, it has been found that neural networks can be easily fooled by adversarial examples, which is a potential safety hazard in some safety-critical applications. Many researchers have proposed various method to make neural networks more robust to white-box adversarial attacks, but an effective method have not been found so far. In this short paper, we focus on the robustness of the features learned by neural networks. We show that the features learned by neural networks are not robust, and find that the robustness of the learned features is closely related to the resistance against adversarial examples of neural networks. We also find that adversarial training against fast gradients sign method (FGSM) does not make the leaned features very robust, even if it can make the trained networks very resistant to FGSM attack. Then we propose a method, which can be seen as an extension of adversarial training, to train neural networks to learn more robust features. We perform experiments on MNIST and CIFAR-10 to evaluate our method, and the experiment results show that this method greatly improves the robustness of the learned features and the resistance to adversarial attacks.

Person re-identification (re-id) suffers from a serious occlusion problem when applied to crowded public places. In this paper, we propose to retrieve a full-body person image by using a person image with occlusions. This differs significantly from the conventional person re-id problem where it is assumed that person images are detected without any occlusion. We thus call this new problem the occluded person re-identitification. To address this new problem, we propose a novel Attention Framework of Person Body (AFPB) based on deep learning, consisting of 1) an Occlusion Simulator (OS) which automatically generates artificial occlusions for full-body person images, and 2) multi-task losses that force the neural network not only to discriminate a person's identity but also to determine whether a sample is from the occluded data distribution or the full-body data distribution. Experiments on a new occluded person re-id dataset and three existing benchmarks modified to include full-body person images and occluded person images show the superiority of the proposed method.

High above the Hudson River in downtown Jersey City, a former U.S. Secret Service agent named Peter Cavicchia III ran special ops for JPMorgan Chase & Co. His insider threat group--most large financial institutions have one--used computer algorithms to monitor the bank's employees, ostensibly to protect against perfidious traders and other miscreants. Aided by as many as 120 "forward-deployed engineers" from the data mining company Palantir Technologies Inc., which JPMorgan engaged in 2009, Cavicchia's group vacuumed up emails and browser histories, GPS locations from company-issued smartphones, printer and download activity, and transcripts of digitally recorded phone conversations. Palantir's software aggregated, searched, sorted, and analyzed these records, surfacing keywords and patterns of behavior that Cavicchia's team had flagged for potential abuse of corporate assets. Palantir's algorithm, for example, alerted the insider threat team when an employee started badging into work later than usual, a sign of potential disgruntlement. That would trigger further scrutiny and possibly physical surveillance after hours by bank security personnel. Over time, however, Cavicchia himself went rogue. Former JPMorgan colleagues describe the environment as Wall Street meets Apocalypse Now, with Cavicchia as Colonel Kurtz, ensconced upriver in his office suite eight floors above the rest of the bank's security team. People in the department were shocked that no one from the bank or Palantir set any real limits.

Many of us now use small doses of AI in everyday life (like Siri, Google Assistant, Alexa, and everything smart home), but hotels are putting this once-sci-fi technology to more widespread use. From concierge robots to personalized rooms to lively chatbots, your next holiday may include help from some artificially intellectualized friends. While you may miss, say, the smile or handshake you get from their human counterparts, these systems can create hyper-personalized experiences and comprehensively upgrade the level of service during your stay. Keep an eye out for these features at your next check-in. Some hotels are using robots to beef up customer service.

The world's first commercial Hyperloop will be running in Abu Dhabi by 2020 and could hurtle passengers around at speeds of up to 760mph (1,200 kmh). The track - which will be around six miles (10km) long - will be close to Al Maktoum International Airport on the border of Abu Dhabi and Dubai. The plan is to have the first section of the track ready for Expo 2020 - which is being held in Dubai - with plans to eventually extend the network to Riyadh, Saudi Arabia, nearly 621 miles (1,000km) away. The world's first commercial Hyperloop will be running in Abu Dhabi by 2020 and could hurtle passengers around at speeds of up to 760mph (1,200 kmh) (concept image) Hilarious moment puppy doesn't want to walk anymore Coronation Street's Rana and Kate caught naked in bed by parents Hilarious moment Billie's daughter Nelly claims she is constipated The track - which will be around six miles (10km) long - will be close to Al Maktoum International Airport on the border of Abu Dhabi and Dubai (concept image). Culver City-based company Hyperloop Transportation Technologies (HyperloopTT) says the super-fast system could also be used to carry cargo from the country's ports and harbours.

Artificial intelligence might be the way of the future -- or increasingly, the present -- but in order to successfully reap the benefits, executives need to ensure that purposeful steps are taken before and during the launch of the software. SAP's report, "Making the Most of Machine Learning: Five Lessons From Fast Learners," reveals the key components of deploying and maximizing machine learning."Machine "Executives need to view machine learning not as a quick fix but as an integral part of a larger strategy to give their business a competitive edge. This requires looking past the initial investment and focusing on the potential for long-term business value."To Half of the participants represented companies with $500 million or more in annual revenue.

The use of artificial intelligence and machine learning in drug development has been ramping up and investors are taking keen notice of the growing trend. This morning London-based BenevolentAI snagged $115 million to give it a total of about $200 million in its coffers. BenevolentAI, which is applying artificial intelligence to develop new medicines for hard to treat diseases, said it will use the funds from the financing round to advance its artificial intelligence driven drug development programs. The company also said it will use the financing to broaden the disease areas on which it focuses, and extend its AI platform capabilities even further. Some of the proceeds from the financing round will be used to explore other science-based industries, including agriculture and energy storage, the company said.

Robots in Singapore have completed a task many humans dread - assembling flat-packed IKEA furniture. Sifting through pages of instructions and a jumble of screws and bolts to build the low-cost Swedish furniture may soon be a thing of the past given advances in technology, say researchers at the city-state's Nanyang Technological University.