The AWS Deep Learning AMIs for Ubuntu and Amazon Linux now come with newer versions of the following deep learning frameworks and interfaces: TensorFlow 1.10 optimized for AWS for higher performance, Horovod 0.13.11 with OpenMPI 3.1.0

Democratized artificial intelligence is one of five trends driving Gartner's latest Hype Cycle for Emerging Technologies, which it updates this month, as it traditionally does during the month of August. The analyst group took a slightly different approach with this year's Hype Cycle for Emerging Tech and grouped the 35 individual technologies into major groups, which includes digitalized ecosystems, do-it-yourself biohacking, transparently immersive experiences, and ubiquitous infrastructure, along with democratized AI. The five trends will "blur the lines between human and machine," Gartner says. "CIOs and technology leaders should always be scanning the market along with assessing and piloting emerging technologies to identify new business opportunities with high impact potential and strategic relevance for their business," writes Mike J. Walker, research vice president at Gartner. The AI group contains technologies like AI platform as a service (PaaS), artificial general intelligence, autonomous driving (levels 4 and 5), autonomous mobile robots, conversational AI platform, deep neural nets, flying autonomous vehicles, smart robots, and virtual assistants.

An AI is set to try and work out how a potentially limitless supply of energy can be used on Earth. It could finally solve the mysteries of fusion power, letting researchers capture and control the process that powers the sun and stars. Researchers at the U.S. Department of Energy's (DOE) Princeton Plasma Physics Laboratory (PPPL) and Princeton University hope to harness a massive new supercomputer to work out how the doughnut-shaped devices, known as tokamaks, can be used. In the middle of the rising Tokamak Building a well is preserved for the ITER machine. While ITER won't generate electricity, scientists hope it will demonstrate that such a fusion reactor can produce more energy than it consumes.

Mumbai: Early this month at the Intel AI Devcon 2018 in Bengaluru, a holographic avatar called Ella listened intently to composer Kevin Doucette playing notes on his synthesizer. When he paused, she began composing her own notes, complementing his music in real-time. Ella was learning about features such as tempo, scale and pitch from the music data that was being sent in real-time to an Intel Movidius Neural Compute Stick. Intel used a class of artificial neural networks, the recurrent neural network or RNN that depends on previous calculations to work on current ones, to perform this artificial intelligence (AI) task. This Neural Compute Stick is simply a case in point that Intel--a company which most people identify with central processing units (CPUs) inside personal computers (PCs), mobiles and servers--is widening its portfolio to stay in the AI race that has strong contenders including Nvidia, Microsoft, Google, Facebook, IBM, Amazon, Apple, Alibaba and Baidu.

2 Diabetes is a leading worldwide public health concern, and its increasing prevalence has significant health and economic importance in all nations. The condition is a multifactorial disorder with a complex aetiology. The genetic determinants remain largely elusive, with only a handful of identified candidate genes. Genome wide association studies (GWAS) promised to significantly enhance our understanding of genetic based determinants of common complex diseases. To date, 83 single nucleotide polymorphisms (SNPs) for type 2 diabetes have been identified using GWAS. Standard statistical tests for single and multi-locus analysis such as logistic regression, have demonstrated little effect in understanding the genetic architecture of complex human diseases. Logistic regression is modelled to capture linear interactions but neglects the non-linear epistatic interactions present within genetic data. There is an urgent need to detect epistatic interactions in complex diseases as this may explain the remaining missing heritability in such diseases. In this paper, we present a novel framework based on deep learning algorithms that deal with non-linear epistatic interactions that exist in genome wide association data. Logistic association analysis under an additive genetic model, adjusted for genomic control inflation factor, is conducted to remove statistically improbable SNPs to minimize computational overheads.

Machine learning has revolutionized fields such as image, text, and speech recognition. There's also growing interest in applying machine and deep learning methods in science, engineering, medicine, and finance. Despite their immense success in practice, there is limited mathematical understanding of neural networks. We mathematically study neural networks in the asymptotic regime of simultaneously (A) large network sizes and (B) large numbers of stochastic gradient descent training iterations. We rigorously prove that the neural network satisfies a central limit theorem. Our result describes the neural network's fluctuations around its mean-field limit. The fluctuations have a Gaussian distribution and satisfy a stochastic partial differential equation.

Recently, considerable effort has been devoted to deep domain adaptation in computer vision and machine learning communities. However, most of existing work only concentrates on learning shared feature representation by minimizing the distribution discrepancy across different domains. Due to the fact that all the domain alignment approaches can only reduce, but not remove the domain shift. Target domain samples distributed near the edge of the clusters, or far from their corresponding class centers are easily to be misclassified by the hyperplane learned from the source domain. To alleviate this issue, we propose to joint domain alignment and discriminative feature learning, which could benefit both domain alignment and final classification. Specifically, an instance-based discriminative feature learning method and a center-based discriminative feature learning method are proposed, both of which guarantee the domain invariant features with better intra-class compactness and inter-class separability. Extensive experiments show that learning the discriminative features in the shared feature space can significantly boost the performance of deep domain adaptation methods.

Extractive summarization models require sentence-level labels, which are usually created heuristically (e.g., with rule-based methods) given that most summarization datasets only have document-summary pairs. Since these labels might be suboptimal, we propose a latent variable extractive model where sentences are viewed as latent variables and sentences with activated variables are used to infer gold summaries. During training the loss comes \emph{directly} from gold summaries. Experiments on the CNN/Dailymail dataset show that our model improves over a strong extractive baseline trained on heuristically approximated labels and also performs competitively to several recent models.

We propose a simple model for textual matching problems. Starting from a Siamese architecture, we augment word embeddings with two features based on exact and paraphrase match between words in the two sentences being considered. We train the model using four types of regularization on datasets for textual entailment, paraphrase detection and semantic relatedness. Our model performs comparably or better than more complex architectures; achieving state-of-the-art results for paraphrase detection on the SICK dataset and for textual entailment on the SNLI dataset.

Character-level features are currently used in different neural network-based natural language processing algorithms. However, little is known about the character-level patterns those models learn. Moreover, models are often compared only quantitatively while a qualitative analysis is missing. In this paper, we investigate which character-level patterns neural networks learn and if those patterns coincide with manually-defined word segmentations and annotations. To that end, we extend the contextual decomposition technique (Murdoch et al. 2018) to convolutional neural networks which allows us to compare convolutional neural networks and bidirectional long short-term memory networks. We evaluate and compare these models for the task of morphological tagging on three morphologically different languages and show that these models implicitly discover understandable linguistic rules. Our implementation can be found at https://github.com/FredericGodin/ContextualDecomposition-NLP .