BENGALURU: The state government on Wednesday announced it will establish a Centre of Excellence for Data Science and Artificial Intelligence (CoE-DS&AI), with Nasscom as the programme and implementation partner. The CoE, being set up at a cost of Rs 40 crore, is said to be first-of-its-kind port based on PPP model, and will accelerate the ecosystem in Karnataka. It will also provide the impetus for development of data science and artificial intelligence across the country. "The state government will further support the initiative with significant share of investment required for initial five years and nonsensitive data which can augment innovation and research in Data Science and Artificial In telligence across the country," the IT & BT department said in a statement. IT and tourism minister Priyank Kharge said the CoE will enable global corporates who plan to set up their global analytics practice in Karnataka and create jobs for 35,000 data science and artificial intelligence professionals over the next five years.

Online giant Amazon, which has invested over Rs13,800 crore in India, is relying heavily on machine learning and artificial intelligence (AI) to goose up market share in the country. Rajeev Rastogi, director for machine learning at Amazon, said both transformative technologies are being used in several ways to improve customer experience. For instance, he said, apart from guiding sellers on categorisation, Amazon also tracks products that have high sales during festivities or holidays, and based on such factors, it suggests the best products and deals that can woo customers. All of these products and services are being built in India, he said. The Seattle-headquartered company soon plans to launch event-based deals in India.

As a technology pioneer, Elon Musk is hardly anyone's idea of a Luddite. So when Musk tweeted that artificial intelligence competition with Russia and China would be the "most likely cause" of World War III, it got people pretty worked up. We were worried about the perils of creating an artificial being long before AI became a science. The term "robot" comes from a play written in 1920 by Karel Čapek called R.U.R -- Rossum's Universal Robots (Rossumovi Univerzální Roboti in the original Czech). The play starts in a factory that manufactures synthetic people to serve as workers.

The Karnataka government on Wednesday announced it will establish a Centre of Excellence for Data Science and Artificial Intelligence (CoE-DS&AI), with Nasscom as the programme and implementation partner. The CoE, being set up at a cost of Rs 40 crore, is said to be first-of-its-kind port based on PPP model, and will accelerate the ecosystem in Karnataka. It will also provide the impetus for development of data science and artificial intelligence across the country. "The state government will further support the initiative with significant share of investment required for initial five years and non sensitive data which can augment innovation and research in Data Science and Artificial Intelligence across the country," the IT & BT department said in a statement. IT and tourism minister Priyank Kharge said the CoE will enable global corpora tes who plan to set up their global analytics practice in Karnataka and create jobs for 35,000 data science and artificial intelligence professionals over the next five years.

STX Entertainment, the Burbank film and television studio behind the comedy "Bad Moms," is moving closer to a planned initial public offering in Hong Kong. The 3-year-old studio is expected to raise about $500 million on the Hong Kong stock exchange early next year, according to a person familiar with the matter. The company, led by Robert Simonds, its chairman and chief executive, has long signaled intentions to raise money from the public market in Hong Kong but financial details and timing have not been revealed. The Wall Street Journal first reported details of the IPO on Wednesday. A representative of STX declined to comment. An IPO by STX would fit a common pattern of companies hoping to capitalize on Chinese investors' interest in entertainment.

It is available from PublicAffairs, an imprint of Perseus Books LLC, a subsidiary of Hachette Book Group Inc. Imagine flying over a major city at night -- say, Chicago or Paris or Beijing -- and it is completely dark below. It is just a void of light akin to nighttime in the middle of the ocean. Then imagine someone flips on the power grid, and you see today's web of human activity light up. Imagine further that someone flips the switch again, and you glimpse a future image of the city. Where you once thought there was nothing, there is a universe of action -- both present and future.

Superstore The staff at Cloud 9 are frantically working to prepare the store for its grand reopening following a tornado's destruction, and Amy and Jonah's (America Ferrera, Ben Feldman) relationship is strained after the intimate moment they shared. Grey's Anatomy The story picks up where the explosive season finale left off as Abigail Spencer joins the cast. Gotham Gordon's (Ben McKenzie) mission is to apprehend Jonathan Crane (guest star Charlie Tahan), but to do so he must return to Arkham. The Vietnam War "The Weight of Memory (March 1973-Onward)," the 10th and final episode of the miniseries by Ken Burns and Lynn Novick, concludes the expansive history of the divisive conflict. The Murder of Laci Peterson: A Closer Look New theories emerge in the wake of the verdict condemning Scott Peterson to death by lethal injection in the series finale of the miniseries.

We consider the problem of clustering noisy finite-length observations of stationary ergodic random processes according to their generative models without prior knowledge of the model statistics and the number of generative models. Two algorithms, both using the $L^1$-distance between estimated power spectral densities (PSDs) as a measure of dissimilarity, are analyzed. The first one, termed nearest neighbor process clustering (NNPC), relies on partitioning the nearest neighbor graph of the observations via spectral clustering. The second algorithm, simply referred to as $k$-means (KM), consists of a single $k$-means iteration with farthest point initialization and was considered before in the literature, albeit with a different dissimilarity measure. We prove that both algorithms succeed with high probability in the presence of noise and missing entries, and even when the generative process PSDs overlap significantly, all provided that the observation length is sufficiently large. Our results quantify the tradeoff between the overlap of the generative process PSDs, the observation length, the fraction of missing entries, and the noise variance. Finally, we provide extensive numerical results for synthetic and real data and find that NNPC outperforms state-of-the-art algorithms in human motion sequence clustering.

The reliable measurement of confidence in classifiers' predictions is very important for many applications and is, therefore, an important part of classifier design. Yet, although deep learning has received tremendous attention in recent years, not much progress has been made in quantifying the prediction confidence of neural network classifiers. Bayesian models offer a mathematically grounded framework to reason about model uncertainty, but usually come with prohibitive computational costs. In this paper we propose a simple, scalable method to achieve a reliable confidence score, based on the data embedding derived from the penultimate layer of the network. We investigate two ways to achieve desirable embeddings, by using either a distance-based loss or Adversarial Training. We then test the benefits of our method when used for classification error prediction, weighting an ensemble of classifiers, and novelty detection. In all tasks we show significant improvement over traditional, commonly used confidence scores.

Adaptive learning rate algorithms such as RMSProp are widely used for training deep neural networks. RMSProp offers efficient training since it uses first order gradients to approximate Hessian-based preconditioning. However, since the first order gradients include noise caused by stochastic optimization, the approximation may be inaccurate. In this paper, we propose a novel adaptive learning rate algorithm called SDProp. Its key idea is effective handling of the noise by preconditioning based on covariance matrix. For various neural networks, our approach is more efficient and effective than RMSProp and its variant.