The development of intelligent traffic light control systems is essential for smart transportation management. While some efforts have been made to optimize the use of individual traffic lights in an isolated way, related studies have largely ignored the fact that the use of multi-intersection traffic lights is spatially influenced and there is a temporal dependency of historical traffic status for current traffic light control. To that end, in this paper, we propose a novel SpatioTemporal Multi-Agent Reinforcement Learning (STMARL) framework for effectively capturing the spatio-temporal dependency of multiple related traffic lights and control these traffic lights in a coordinating way. Specifically, we first construct the traffic light adjacency graph based on the spatial structure among traffic lights. Then, historical traffic records will be integrated with current traffic status via Recurrent Neural Network structure. Moreover, based on the temporally-dependent traffic information, we design a Graph Neural Network based model to represent relationships among multiple traffic lights, and the decision for each traffic light will be made in a distributed way by the deep Q-learning method. Finally, the experimental results on both synthetic and real-world data have demonstrated the effectiveness of our STMARL framework, which also provides an insightful understanding of the influence mechanism among multi-intersection traffic lights.

Machine comprehension, answering a question depending on a given context paragraph is a typical task of Natural Language Understanding. It requires to model complex dependencies existing between the question and the context paragraph. There are many neural network models attempting to solve the problem of question answering. The best models have been selected, studied and compared with each other. All the selected models are based on the neural attention mechanism concept. Additionally, studies on a SQUAD dataset were performed. The subsets of queries were extracted and then each model was analyzed how it deals with specific group of queries. Based on these three model ensemble model was created and tested on SQUAD dataset. It outperforms the best Mnemonic Reader model.

It would be difficult nowadays not to come across a piece of news, article or posting related to Artificial Intelligence, or AI. Whether it is smart appliances and home automation systems with intelligent digital assistants, autonomous vehicles, advanced medical diagnostics, virtual reality, or even human-like automatons supplanting humans themselves – such as the recent news of the humanoid robot piloting a spacecraft and attempting to dock at the International Space Station – there seems to be no stopping AI. AI is not a new concept. In fact, it was coined over sixty years ago by John McCarthy during a conference with fellow scientists at Dartmouth College in New Hampshire. Yet AI, described as the ability for machines to simulate intellectual processes characteristic of humans, is an area of study that has existed long before its inception as an academic discipline.

Nvidia announced its new enterprise software product, vComputeServer, which has been developed and optimised for use with VMware's vSphere. Last week, VMware announced its intention to acquire Carbon Black and Pivotal, in a massive deal that will expand the company's SaaS offerings, while enhancing its ability to enable digital transformation for customers. Before the dust had even settled on that news, the company announced today (26 August), that it is set to launch a hybrid cloud on AWS (Amazon Web Services) in partnership with Nvidia, which will improve GPU (graphics processing unit) virtualisation. The two companies say that this is the first hybrid cloud service that lets enterprises accelerate AI, machine learning or deep learning workloads with GPUs. At the VMWorld conference in San Francisco, Nvidia's VP of product management, John Fanelli, told reporters: "In a modern data centre, organisations are going to be using GPUs to power AI, deep learning and analytics. "Due to the scale of those types of workloads, they're going to be doing some processing on premise in data centres, some processing in clouds and continually iterating between them." The company said that this will make the completion of deep learning training up to 50 times faster than with a CPU alone. This product is aimed at people who may be using Nvidia's Rapids software, Fanelli explained, which is a suite of data processing and machine learning libraries used for GPU-acceleration in data science workflows. Nvidia founder and CEO Jensen Huang said: "From operational intelligence to artificial intelligence, businesses rely on GPU-accelerated computing to make fast, accurate predictions that directly impact their bottom line.

Many vendors like Cisco and Intel are proponents of Edge Processing (also called Edge computing). The main idea behind Edge Computing is to push processing away from the core and towards the Edge of the network. For IoT, that means pushing processing towards the sensors or a gateway. This enables data to be initially processed at the Edge device possibly enabling smaller datasets sent to the core. Devices at the Edge may not be continuously connected to the network.

Many vendors like Cisco and Intel are proponents of Edge Processing (also called Edge computing). The main idea behind Edge Computing is to push processing away from the core and towards the Edge of the network. For IoT, that means pushing processing towards the sensors or a gateway. This enables data to be initially processed at the Edge device possibly enabling smaller datasets sent to the core. Devices at the Edge may not be continuously connected to the network.

What does Google know that the other 99.99% of organizations don't? I don't roam the hallowed hallways of Google or have access to any insider secrets as to their business plans and future business models, but it sure does make one contemplate why they would give away the very engine that fuels their obscenely-lucrative search business. You see, Google uses Machine Learning and Deep Learning to fuel its Personal Photo App (it auto-magically groups your photos into storyboards or collages), recognize spoken words (natural language processing), translate foreign languages (handy for me as I frequently travel internationally), and serve as the basis for its search engine. Arguably, Machine Learning and Deep Learning are the foundation for everything that makes Google money, and TensorFlow is the foundation for those Machine Learning and Deep Learning efforts. So why would Google open source TensorFlow and make it accessible to everyone – researchers, scientists, machine learning experts, students, and even its competitors?

This review presents machine learning (ML) approaches from an applied economist's perspective. We first introduce the key ML methods drawing connections to econometric practice. We then identify current limitations of the econometric and simulation model toolbox in applied economics and explore potential solutions afforded by ML. We dive into cases such as inflexible functional forms, unstructured data sources and large numbers of explanatory variables in both prediction and causal analysis, and highlight the challenges of complex simulation models. Finally, we argue that economists have a vital role in addressing the shortcomings of ML when used for quantitative economic analysis. Machine learning (ML) offers great potential for expanding the applied economist's toolbox. ML tools are beginning to be employed in economic analysis (März et al., 2016; Crane-Droesch, 2017; Athey, 2019), while some researchers raise concerns about their transparency, interpretability and use for ...

The Sephora Studio offers the ultimate makeup experience--and more are opening in 2018. Amanda Li, 25, looked at herself in a mirror. She saw her own reflection, but with extra blush, extended eyelashes and a different lip color. As she blinked, her look suddenly changed, this time with a smoky eyeliner. The scene was part of Sephora's efforts called Virtual Artist, which debuted in 2016 and has gotten a makeover over time.

The results appear in a paper published online on August 19 in the journal Nature Photonics. "An historic issue with OCT is that the depth resolution is typically several times better than the lateral resolution," said Joseph Izatt, the Michael J. Fitzpatrick Professor of Engineering at Duke. "If the layers of imaged tissues happen to be horizontal, then they're well defined in the scan. But to extend the full power of OCT for live imaging of tissues throughout the body, a method for overcoming the tradeoff between lateral resolution and depth of imaging was needed." OCT is an imaging technology analogous to ultrasound that uses light rather than soundwaves. A probe shoots a beam of light into a tissue and, based on the delays of the light waves as they bounce back, determines the boundaries of the features within.