Japan plans to build the world's fastest supercomputer in a bid to arm its manufacturers with a platform for research that could help them develop and improve driverless cars, robotics and medical diagnostics. The Ministry of Economy, Trade and Industry will spend ¥19.5 billion ($173 million) on the previously unreported project, a budget breakdown shows, as part of a government policy to get back Japan's mojo in the world of technology. The country has lost its edge in many electronic fields amid intensifying competition from South Korea and China, which is home to the world's current best-performing machine. In a move that is expected to vault Japan to the top of the supercomputing heap, its engineers will be tasked with building a machine that can make 130 quadrillion calculations per second -- or 130 petaflops in scientific parlance -- as early as next year, sources involved in the project said. At that speed, Japan's computer would be ahead of China's Sunway Taihulight, which is capable of 93 petaflops.

DeepMind, the British AI firm owned by Google's parent company Alphabet, has signed a new five-year deal to use data collected by the UK's National Health Service. The agreement with the NHS Royal Free Hospital Trust in London replaces a previous deal that attracted controversy over its lack of official oversight. Under the terms of the new deal, DeepMind will handle personally identifiable medical records for some 1.6 million patients, including medical history dating back five years. The agreement also includes stricter data regulation, including "technical audits" of DeepMind's systems. Using data from the Royal Free, DeepMind has built an app named Streams that alerts doctors when patients are in danger of developing acute kidney injury (AKI) -- a common but often overlooked condition.

On January 2016, the White House announced its aim to deliver a decade's worth of advances in cancer prevention, diagnosis, and treatment, in five years with this initiative. This initiative includes the building of an AI framework named CANDLE (Cancer Distributed Learning Environment) being developed by multiple organizations. It will help us change the way we understand cancer. READ MORE 6. "GPU Deep Learning has given us a new tool to tackle grand challenges that have, up to now, been too complex for even the most powerful supercomputers. Together with the Department of Energy and the National Cancer Institute we are creating an AI supercomputing platform for cancer research."

According to Yann LeCun, "adversarial training is the coolest thing since sliced bread". I'm inclined to believe so because I don't think sliced bread ever created this much buzz and excitement within the deep learning community. In this post, we'll be looking at 3 papers that built on the pioneering work of Ian Goodfellow in 2014.

Google's Machine Learning framework TensorFlow was open-sourced in November 2015 [1] and has since built a growing community around it. TensorFlow is supposed to be flexible for research purposes while also allowing its models to be deployed productively. This work is aimed towards people with experience in Machine Learning considering whether they should use TensorFlow in their environment. Several aspects of the framework important for such a decision are examined, such as the heterogenity, extensibility and its computation graph. A pure Python implementation of linear classification is compared with an implementation utilizing TensorFlow. I also contrast TensorFlow to other popular frameworks with respect to modeling capability, deployment and performance and give a brief description of the current adaption of the framework.

Modern automatic speech recognition (ASR) systems need to be robust under acoustic variability arising from environmental, speaker, channel, and recording conditions. Ensuring such robustness to variability is a challenge in modern day neural network-based ASR systems, especially when all types of variability are not seen during training. We attempt to address this problem by encouraging the neural network acoustic model to learn invariant feature representations. We use ideas from recent research on image generation using Generative Adversarial Networks and domain adaptation ideas extending adversarial gradient-based training. A recent work from Ganin et al. proposes to use adversarial training for image domain adaptation by using an intermediate representation from the main target classification network to deteriorate the domain classifier performance through a separate neural network. Our work focuses on investigating neural architectures which produce representations invariant to noise conditions for ASR. We evaluate the proposed architecture on the Aurora-4 task, a popular benchmark for noise robust ASR. We show that our method generalizes better than the standard multi-condition training especially when only a few noise categories are seen during training.

Recently exciting progress has been made on protein contact prediction, but the predicted contacts for proteins without many sequence homologs is still of low quality and not very useful for de novo structure prediction. This paper presents a new deep learning method that predicts contacts by integrating both evolutionary coupling (EC) and sequence conservation information through an ultra-deep neural network formed by two deep residual networks. This deep neural network allows us to model very complex sequence-contact relationship as well as long-range inter-contact correlation. Our method greatly outperforms existing contact prediction methods and leads to much more accurate contact-assisted protein folding. Tested on three datasets of 579 proteins, the average top L long-range prediction accuracy obtained our method, the representative EC method CCMpred and the CASP11 winner MetaPSICOV is 0.47, 0.21 and 0.30, respectively; the average top L/10 long-range accuracy of our method, CCMpred and MetaPSICOV is 0.77, 0.47 and 0.59, respectively. Ab initio folding using our predicted contacts as restraints can yield correct folds (i.e., TMscore>0.6) for 203 test proteins, while that using MetaPSICOV- and CCMpred-predicted contacts can do so for only 79 and 62 proteins, respectively. Further, our contact-assisted models have much better quality than template-based models. Using our predicted contacts as restraints, we can (ab initio) fold 208 of the 398 membrane proteins with TMscore>0.5. By contrast, when the training proteins of our method are used as templates, homology modeling can only do so for 10 of them. One interesting finding is that even if we do not train our prediction models with any membrane proteins, our method works very well on membrane protein prediction. Finally, in recent blind CAMEO benchmark our method successfully folded 5 test proteins with a novel fold.

The next time you enter a query into Google's search engine or consult the company's map service for directions to a movie theater, remember that a big brain is working behind the scenes to provide relevant search results and make sure you don't get lost while driving. As Fortune's Roger Parloff wrote, the Google Brain research team has created over 1,000 so-called deep learning projects that have supercharged many of Google's products over the past few years like YouTube, translation, and photos. With deep learning, researchers can feed huge amounts of data into software systems called neural nets that learn to recognize patterns within the vast information faster than humans. In an interview with Fortune, one of Google Brain's co-founders and leaders, Jeff Dean, talks about cutting-edge A.I. research, the challenges involved, and using A.I. in its products. The following has been edited for length and clarity. A lot of human learning comes from unsupervised learning where you're just sort of observing the world around you and understanding how things behave.

Intel Corp. (NASDAQ:INTC) is set on dominating markets that are bound to become massive cash mines in the future. There is significant growth expected from its Data Center business. The company recently announced it would dedicate $250 million to the development of self-driving car systems. The chip maker is also gunning to secure a stronghold in the development of Artificial Intelligence (AI) technology. Analysts are confident about Intel's capacity to rake in the IoT segment in the process.

A few short months ago, Intel acquired Nervana Systems for 400 million dollars with the intention of using the technology they developed in order to be competitive in the deep learning market currently dominated by GPU-based solutions from NVIDIA. Artificial Intelligence is a big market for Intel and the company sees it as a pivotal ground that they must put a stake in or risk falling behind like they did on the mobile front. With Nervana's technology, Intel is expecting to produce "a breakthrough 100-fold increase in performance in the next three years to train complex neural networks", says Intel CEO Bryan Krzanich in a recent editorial. Nervana's technology will be a PCIe add-in card expected to hit be out sometime around the first half of 2017, codenamed Lake Crest and incorporates HBM technology that is directly targeting current GPU solutions. Intel believes that GPGPU architecture is not uniquely advantageous for AI and that their approach can support much larger models and is much more highly scalable.