Similarity/Distance measures play a key role in many machine learning, pattern recognition, and data mining algorithms, which leads to the emergence of metric learning field. Many metric learning algorithms learn a global distance function from data that satisfy the constraints of the problem. However, in many real-world datasets that the discrimination power of features varies in the different regions of input space, a global metric is often unable to capture the complexity of the task. To address this challenge, local metric learning methods are proposed that learn multiple metrics across the different regions of input space. Some advantages of these methods are high flexibility and the ability to learn a nonlinear mapping but typically achieves at the expense of higher time requirement and overfitting problem. To overcome these challenges, this research presents an online multiple metric learning framework. Each metric in the proposed framework is composed of a global and a local component learned simultaneously. Adding a global component to a local metric efficiently reduce the problem of overfitting. The proposed framework is also scalable with both sample size and the dimension of input data. To the best of our knowledge, this is the first local online similarity/distance learning framework based on PA (Passive/Aggressive). In addition, for scalability with the dimension of input data, DRP (Dual Random Projection) is extended for local online learning in the present work. It enables our methods to be run efficiently on high-dimensional datasets, while maintains their predictive performance. The proposed framework provides a straightforward local extension to any global online similarity/distance learning algorithm based on PA.

We conduct a large-scale study of language models for chord prediction. Specifically, we compare N-gram models to various flavours of recurrent neural networks on a comprehensive dataset comprising all publicly available datasets of annotated chords known to us. This large amount of data allows us to systematically explore hyper-parameter settings for the recurrent neural networks---a crucial step in achieving good results with this model class. Our results show not only a quantitative difference between the models, but also a qualitative one: in contrast to static N-gram models, certain RNN configurations adapt to the songs at test time. This finding constitutes a further step towards the development of chord recognition systems that are more aware of local musical context than what was previously possible.

In recent times, the manufacturing processes are faced with many external or internal (the increase of customized product rescheduling , process reliability,..) changes. Therefore, monitoring and quality management activities for these manufacturing processes are difficult. Thus, the managers need more proactive approaches to deal with this variability. In this study, a proactive quality monitoring and control approach based on classifiers to predict defect occurrences and provide optimal values for factors critical to the quality processes is proposed. In a previous work (Noyel et al. 2013), the classification approach had been used in order to improve the quality of a lacquering process at a company plant; the results obtained are promising, but the accuracy of the classification model used needs to be improved. One way to achieve this is to construct a committee of classifiers (referred to as an ensemble) to obtain a better predictive model than its constituent models. However, the selection of the best classification methods and the construction of the final ensemble still poses a challenging issue. In this study, we focus and analyze the impact of the choice of classifier types on the accuracy of the classifier ensemble; in addition, we explore the effects of the selection criterion and fusion process on the ensemble accuracy as well. Several fusion scenarios were tested and compared based on a real-world case. Our results show that using an ensemble classification leads to an increase in the accuracy of the classifier models. Consequently, the monitoring and control of the considered real-world case can be improved.

Recently, the online car-hailing service, Didi, has emerged as a leader in the sharing economy. Used by passengers and drivers extensive, it becomes increasingly important for the car-hailing service providers to minimize the waiting time of passengers and optimize the vehicle utilization, thus to improve the overall user experience. Therefore, the supply-demand estimation is an indispensable ingredient of an efficient online car-hailing service. To improve the accuracy of the estimation results, we analyze the implicit relationships between the points of Interest (POI) and the supply-demand gap in this paper. The different categories of POIs have positive or negative effects on the estimation, we propose a POI selection scheme and incorporate it into XGBoost [1] to achieve more accurate estimation results. Our experiment demonstrates our method provides more accurate estimation results and more stable estimation results than the existing methods.

Amagi, a global leader in cloud-based technology for media processing, today announced the launch of TORNADO, a machine learning-based content preparation service that enables TV networks and content owners to scale their operations, accelerate broadcast workflows, generate new revenues and reduce operational costs. Compared to traditional manual content preparation, Amagi TORNADO is nearly six times more efficient, allowing broadcasters free up capital and streamline workflows. Over the last three years, the broadcast industry has had to evolve significantly due to a rise in multi-screen content consumption, demands for "here and now" content and a shift in how consumers are viewing content as more consumers move from cable to OTT services. In such an evolving scenario, TV networks, content owners, and digital-first networks are creatively trying to grab a piece of the action by trying new mediums and delivery methods to provide better experiences to consumers while streamlining costs and operations. However, despite these efforts, content preparation continues to require pain-staking hours of manual work and massive overhead costs.

The term'financial crimes' often brings to mind issues such as fraud, lottery scams or credit card skimming, but today there are far-reaching crimes that many fail to consider. This includes money laundering, human and drug trafficking, the financing of terrorism, and the bribery of public officials, all of which are much broader financial crimes that impact organizations, governments and even individuals. The sinister and disruptive nature of these crimes affect the most fragile economies at a high level - and the most vulnerable persons at the lowest level. It clearly impacts the reputation of countries and how their citizens are viewed internationally. But worst of all, it reduces the sense of natural justice that most people feel when crimes such as these escape detection and punishment.

NEW YORK – A SpaceX capsule carrying food, experiments and other goods for NASA has arrived at the International Space Station after a two-day journey. The Dragon capsule and its 6,000-pound shipment was captured by the space station's robot arm Wednesday. It's the second trip to the 250-mile-high orbiting outpost for this capsule, refurbished following a visit two years ago. It will remain attached to the space station for about a month, returning to Earth in May. The space station is currently home to astronauts from the U.S., Russia and Japan. The supply capsule launched Monday from Cape Canaveral, Florida, aboard a used Falcon rocket.

Most AI assistants can't really hold a conversation. They're fine with I-go-you-go dialogue, but most humans aren't quite so timid -- they know when to interrupt, and when to restart chat when there's an awkward pause. Microsoft wants to fix that. It just upgraded its Xiaolce chatbot AI with "full duplex" conversation that lets it start speaking when it's listening to what you're saying. As it can predict what you're likely to say next, it knows when to interrupt you with important info or say something more when both sides suddenly go quiet.

In the United States, there are over 5 million young adults between the ages of 18-35 living alone, and that number is growing. While many of them may be living alone by choice, it can also be socially isolating, if you're into that whole being social thing. The situation is similar in many other countries, especially in Asia. There are plenty of robots under development (and even available) for elderly people with social isolation issues, but younger people are expected to, uh, just go outside or something. At the ACM/IEEE International Conference on Human Robot Interaction last month, roboticists from Korea introduced a robot called Fribo, which is designed to provide a way for young adults who live alone to maintain daily connections with one another.

Apple (AAPL) aims to catch-up in artificial intelligence as it has hired away the former head of AI and search technology from Google-parent Alphabet (GOOGL). John Giannandrea will report directly to Apple Chief Executive Tim Cook, according to the New York Times. Apple's move is the latest example of tech giants luring away top AI talent. The companies are rushing to improve a wide range of existing products with AI tools. AI programs -- essentially computer algorithms -- analyze huge amounts of data to identify patterns and predict outcomes.