The task of entity relation extraction discovers new relation facts and enables broader applications of knowledge graph. Distant supervision is widely adopted for relation extraction, which requires large amounts of texts containing entity pairs as training data. However, in some specific domains such as medicalrelated applications,entity pairs that have certain relations might not appear together, thus it is difficult to meet the requirement for distantly supervised relation extraction. In the light of this challenge, we propose a novel path-based model to discover new entity relation facts. Instead of finding texts for relation extraction, the proposed method extracts path-only information for entity pairs from the current knowledgegraph. For each pair of entities, multiple paths can be extracted, and some of them are more useful for relation extraction than others. In order to capture this observation, we employ attention mechanism to assign different weights for different paths, which highlights the useful paths for entity relation extraction. To demonstrate the effectiveness of the proposed method, we conduct various experiments on a large-scale medical knowledge graph. Compared with the state-of-the-art relation extraction methods using the structure of knowledge graph, the proposed method significantly improves the accuracy of extracted relation factsand achieves the best performance.

In this paper, we explore the utilization of natural language to drive transfer for reinforcement learning (RL). Despite the wide-spread application of deep RL techniques, learning generalized policy representations that work across domains remains a challenging problem. We demonstrate that textual descriptions of environments provide a compact intermediate channel to facilitate effective policy transfer. Specifically, by learning to ground the meaning of text to the dynamics of the environment such as transitions and rewards, an autonomous agent can effectively bootstrap policy learning on a new domain given its description. We employ a model-based RL approach consisting of a differentiable planning module, a model-free component and a factorized state representation to effectively use entity descriptions. Our model outperforms prior work on both transfer and multi-task scenarios in a variety of different environments. For instance, we achieve up to 14% and 11.5% absolute improvement over previously existing models in terms of average and initial rewards, respectively.

This article looks at the ways in which firms across the various sectors of the economy adopt Artificial Intelligence (AI) techniques. However, before we review the sectors affected it is important to note the underlying drivers that are fuelling the growth in the influence and reach of Machine Learning across the sectors of the economy will only grow as we move forwards. This is because Big Data is only getting larger, velocity of data faster, plus the availability of cheaper data storage plus the arrival of powerful Graphical Processing Units (GPUs) to enable Deep Learning algorithms to be deployed. Furthermore, new research in areas of Deep Learning and other Machine Learning areas will continue to emerge into real world production over the next few years leading to new opportunities and applications. The DLS team strongly believe that the advent of 5G around 2021 will be a transformative and revolutionary moment in human history.

IBM is unveiling new hardware that brings power efficiency and improved training times to artificial intelligence (AI) projects this week at the International Electron Devices Meeting (IEDM) and the Conference on Neural Information Processing Systems (NeurIPS), with 8-bit precision for both their analog and digital chips for AI. Over the last decade, computing performance for AI has improved at a rate of 2.5x per year, due in part to the use of GPUs to accelerate deep learning tasks, the company noted in a press release. However, this improvement is not sustainable, as most of the potential performance from this design model--a general-purpose computing solution tailored to AI--will not be able to keep pace with hardware designed exclusively for AI training and development. Per the press release, "Scaling AI with new hardware solutions is part of a wider effort at IBM Research to move from narrow AI, often used to solve specific, well-defined tasks, to broad AI, which reaches across disciplines to help humans solve our most pressing problems." While traditional computing has been in a decades-long path of increasing address width--with most consumer, professional, and enterprise-grade hardware using 64-bit processors--AI is going the opposite direction.

Shervine Amidi, graduate student at Stanford, and Afshine Amidi, of MIT and Uber -- creators of a recent set of machine leanring cheat sheets -- have just published a new set of deep learning cheat sheets. These "VIP cheat sheets" are based on the materials from Stanford's CS 230 (Github repo with PDFs available here), and include topics such as: You can visit Shervine's Stanford resource page or the Github repo for more information, or can download the cheat sheets from the direct download links above. You can also find all of the sheets bundled together into a single "super VIP cheat sheet."

Artificial intelligence (AI) is rapidly transforming everything today, from daily lives to transportation to businesses. Humans have always found the concept of AI very enthralling as is evident from the number of hit sci-fi movies. Scientists and researchers have worked hard on making this technology a norm for the human beings. Enterprises are adopting AI and machine learning (ML) for various use cases, which has risen the demand for AI engines that can be used to develop intelligent applications and tools. Such apps and tools help them automate the repetitive, tedious and difficult tasks that can affect productivity and cost of operation.

Monday, December 3, Rockville, MD - Insilico Medicine, one of the leaders in artificial intelligence for drug discovery, biomarker development, digital medicine, and aging research announced today the publication of its recent paper titled "Artificial Intelligence for Aging and Longevity Research: Recent Advances and Perspectives" in Ageing Research Reviews, one of the highest-impact journals in the field. The paper introduces recent advances in deep learning for aging research and provides fair insight into the field. The emergence of the longevity biotechnology industry has brought many biotech and pharma companies and academic research institutions to the longevity landscape, and now one of the key trends accelerating the field is recent advances in artificial intelligence. "Insilico Medicine is dedicated to extending human longevity. We came up with several very important realizations. First, age is one of the most abundant biological features, and when your data looks like Swiss cheese, age is present. Second, Deep Learning (DL) age predictors are a great way to integrate previously incompatible data types, such as videos and blood test results. Third, the generation of new biological data using Generative Adversarial Networks (GANs), with age as a generation condition, is a great way to produce high-quality synthetic data. Also, it is possible to view aging as a staged disease to get a holistic view of the biological process on both tissue-specific and systemic levels, which makes the Deep Neural Networks (DNNs) more interpretable, builds causal graphs, and identifies biological targets. Moreover, it is possible to train the DNNs on age and retrain the model on specific diseases. Also, it is possible to use biological aging clocks to personalize immunotherapies and vaccinations and to identify new ways to improve response rates. The paper outlines these realizations and presents a way to accelerate aging research using AI technologies", said Alex Zhavoronkov, Ph.D., founder and CEO of Insilico Medicine, who led the study.

Artificial Intelligence (AI) is an intriguing concept that has fascinated experts and laymen alike for years now. Technology in 2018 is moving at a breakneck speed, and it is safe to say that man today has significantly more power in his pocket than he had in his entire home back in the 90s. There have been immense breakthroughs in the field of machine learning and deep learning. These concepts have allowed machines to process and analyze information themselves in a very sophisticated manner. Thanks to these AI developments; machines can now perform complex functions such as facial recognition.

There are some detection problems in the world that only experts can solve, and by doing so are saving lives every day. Radiologists looking for intracerebral hemorrhage (ICH) save lives, but their time is scarce and expensive. But what if we could build an AI to perform this sort of detection? It is no simple task to train a CNN model, such as U-Net, to achieve this. But with the progress of deep learning libraries such as TensorFlow, the revolution of cloud providers such as AWS, Azure, and GCP, and deep learning platforms such as MissingLink, it's becoming increasingly feasible for startups to build an app at almost any scale--including to mimic the work of radiologists and other experts.

Humans gain an implicit understanding of physical laws through observing and interacting with the world. Endowing an autonomous agent with an understanding of physical laws through experience and observation is seldom practical: we should seek alternatives. Fortunately, many of the laws of behaviour of the physical world can be derived from prior knowledge of dynamical systems, expressed through the use of partial differential equations. In this work, we suggest a neural network capable of understanding a specific physical phenomenon: wave propagation in a two-dimensional medium. We define `understanding' in this context as the ability to predict the future evolution of the spatial patterns of rendered wave amplitude from a relatively small set of initial observations. The inherent complexity of the wave equations -- together with the existence of reflections and interference -- makes the prediction problem non-trivial. A network capable of making approximate predictions also unlocks the opportunity to speed-up numerical simulations for wave propagation. To this aim, we created a novel dataset of simulated wave motion and built a predictive deep neural network comprising of three main blocks: an encoder, a propagator made by 3 LSTMs, and a decoder. Results show reasonable predictions for as long as 80 time steps into the future on a dataset not seen during training. Furthermore, the network is able to generalize to an initial condition that is qualitatively different from those seen during training.