The Movidius Neural Compute Stick enables rapid prototyping, validation and deployment of Deep Neural Network (DNN) inference applications at the edge. Its low-power VPU architecture enables an entirely new segment of AI applications that aren't reliant on a connection to the cloud. The NCS combined with Movidius Neural Compute SDK allows deep learning developers to profile, tune, and deploy Convolutional Neural Network (CNN) on low-power applications that require real-time inferencing.

Identifying and pinpointing potential network failures and performance issues has long been a matter of educated guesswork, but an emerging generation of predictive analytics tools promises to bring greater accuracy to network reliably forecasts, allowing staff to address and remedy specific issues even before they can even begin affecting network operations. Predictive analytics is a game-changer, giving CIOs the ability to literally look into the future. "There is a growing need for networks to adapt to dynamic application demands as well as address dynamically to special events, seasonality and so on," says Diomedes Kastanis, head of technology and innovation for Ericsson. "Although we have a lot of automation systems and rules to manage and operate networks, it still it not enough to cope with the intense changing environment and proactively adapt to changing demands." Predictive analytics incorporating processes such as machine learning (ML) and artificial intelligence (AI) are relatively new concepts to many CIOs.

Deep learning is a new method of artificial intelligence that is an active, fast-moving area of research where we can expect advances to become market-ready over the next several years. Unfortunately, market hype has turned deep learning into a buzzword that can contribute to the misconception that other approaches to AI are not relevant. After all, if you are not doing deep learning, surely you must be doing shallow learning, right? In cybersecurity, we use various techniques, such as statistics, probability theory, and multiple machine learning algorithms (of which deep learning is one example), to look at use cases and the data available, selecting the best math or algorithm for the job. We take data from various sources -- application logs, source code, etc. -- choosing the right algorithms based on our understanding of the dataset and use case.

The engineer in charge of keeping Google available 24/7 says the enormous datacenters that underpin online services worldwide will soon be run with the help of AI. Ben Treynor Sloss, Google's VP of engineering, is basing this prediction on the profound savings identified by a machine-learning system charged with helping run a Google datacenter in 2016. The Google DeepMind system significantly improved the power efficiency of the Google datacenter, via tweaks to how servers were run and the operation of power and cooling equipment. Following the system's advice allowed Google to reduce the energy needed to cool servers by about 40 percent. If that reduction were replicated across Google's datacenters worldwide, it could add up to a saving of tens of millions of dollars each year.

We propose a novel decoding approach for neural machine translation (NMT) based on continuous optimisation. We convert decoding - basically a discrete optimization problem - into a continuous optimization problem. The resulting constrained continuous optimisation problem is then tackled using gradient-based methods. Our powerful decoding framework enables decoding intractable models such as the intersection of left-to-right and right-to-left (bidirectional) as well as source-to-target and target-to-source (bilingual) NMT models. Our empirical results show that our decoding framework is effective, and leads to substantial improvements in translations generated from the intersected models where the typical greedy or beam search is not feasible. We also compare our framework against reranking, and analyse its advantages and disadvantages.

Find out how to optimize your website to give your customers experiences that will have the biggest ROI for your business. Machine learning isn't new to the enterprise, but technological advances and accelerating investments have made it available to the average entrepreneur. In fact, according to a recent Forrester survey, machine-learning investments are increasing 300 percent this year compared to last year. Already, machine learning has made its mark in areas like self-driving cars, personalized-content recommendations and even face-recognition filters. Clearly, though, machine learning can do much more than suggest content and steer cars.

Microsoft is building an artificial intelligence (AI) research hub to speed up the integration of AI into products and services. Harry Shum, executive vice president for Microsoft's AI and research group, spoke about the new lab, called Microsoft Research AI, for the first time this week at an event in the U.K. In a blog post, he wrote that the lab will combine various disciplines -- such as machine learning, perception, and natural language processing -- to develop more sophisticated AI. This integrated approach aims to develop systems that can understand language and take action based on that understanding. Machine reading, which combines AI disciplines such as natural language processing and deep learning, is an example. "We believe AI will be even more helpful when we can create tools that combine those functions and add some of the abilities that come naturally to people," Shum wrote.

Today, Intel launched the Movidius Neural Compute Stick, the world's first USB-based deep learning inference kit and self-contained artificial intelligence (AI) accelerator that delivers dedicated deep neural network processing capabilities to a wide range of host devices at the edge. Designed for product developers, researchers and makers, the Movidius Neural Compute Stick aims to reduce barriers to developing, tuning and deploying AI applications by delivering dedicated high-performance deep-neural network processing in a small form factor. As more developers adopt advanced machine learning approaches to build innovative applications and solutions, Intel is committed to providing the most comprehensive set of development tools and resources to ensure developers are retooling for an AI-centric digital economy. Whether it is training artificial neural networks on the Intel Nervana cloud, optimizing for emerging workloads such as artificial intelligence, virtual and augmented reality, and automated driving with Intel Xeon Scalable processors, or taking AI to the edge with Movidius vision processing unit (VPU) technology, Intel offers a comprehensive AI portfolio of tools, training and deployment options for the next generation of AI-powered products and services. "The Myriad 2 VPU housed inside the Movidius Neural Compute Stick provides powerful, yet efficient performance – more than 100 gigaflops of performance within a 1W power envelope – to run real-time deep neural networks directly from the device," said Remi El-Ouazzane, vice president and general manager of Movidius, an Intel company.

The Intelligence Advanced Research Projects Activity (IARPA) recently opened registration for a prize challenge in automated analysis and classification of objects within satellite imagery. IARPA's Functional Map of the World (fMoW) Challenge asks developers to build machine learning algorithms capable of classifying the function of facilities, specific buildings, and land--information used by intelligence personnel to support defense, humanitarian, and disaster response missions. "We are hoping to introduce learning opportunities for the geospatial and deep learning communities to integrate their approaches and increase the exposure to the scientific gains that could be made by combining these two disciplines," said IARPA spokesperson Charles Carithers. Labeling objects within satellite imagery is time-consuming when handled by a human analyst and contributes to operator burnout. This challenge aims to produce breakthroughs in deep learning analysis that will accelerate this process and, in the absence of human error, improve accuracy.

Reinforcement learning algorithms that can reliably learn how to control robots, etc. Better generative models. Algorithms that can reliably learn how to generate images, speech and text that humans can't tell apart from the real thing. Learning to learn and ubiquitous deep learning. Right now it still takes a human expert to run the learning-to-learn algorithm, but in the future it will be easier to deploy, and all kinds of businesses that don't specialize in AI will be able to leverage deep learning. More cyberattacks will leverage machine learning to make more autonomous malware, more efficient fuzzing for vulnerabilities, etc. More cyberdefenses will leverage machine learning to respond faster than a human could, detect more subtle intrusions, etc. ML algorithms from opposing camps will fool each other to carry out both attacks and defensive actions.