An AI has learned to tackle one of the toughest Atari video games by taking instructions in plain English. The system, developed by a team at Stanford University in California, learned to play the game Montezuma's Revenge, in which players scour an Aztec temple for treasure. The game is challenging for AI to learn because it offers sparse rewards, requiring players to make several moves before earning any points. Most video-game-playing AIs use reinforcement learning to develop a strategy, relying on feedback like game points to tell them when they are playing well. To help their AI pick up game tactics quicker, the Stanford team gave their reinforcement learning system a helping hand in the form of natural language instructions, for example advising it to "climb up the ladder" or "get the key".

Almost a year ago QuantStart discussed deep learning and introduced the Theano library via a logistic regression example. Given the recent results of the QuantStart 2017 Content Survey it was decided that an up to date beginner-friendly article was needed to introduce deep learning from first principles. These days it is almost impossible to work in any technology-heavy field without hearing about the latest advances in the field of deep learning. Quantitative finance is no different. Many of the recent discussions in the latest quant finance conferences such as Quantopian's QuantCon and Newsweek's AI & Data Science - Capital Markets are largely focusing around the promise of deep learning as the next frontier in quantitative trading.

Google's TensorFlow has been a hot topic in deep learning recently. The open source software, designed to allow efficient computation of data flow graphs, is especially suited to deep learning tasks. It is designed to be executed on single or multiple CPUs and GPUs, making it a good option for complex deep learning tasks. In it's most recent incarnation – version 1.0 – it can even be run on certain mobile operating systems. This introductory tutorial to TensorFlow will give an overview of some of the basic concepts of TensorFlow in Python. These will be a good stepping stone to building more complex deep learning networks, such as Convolution Neural Networks and Recurrent Neural Networks, in the package.

Babylon Health, the U.K. startup that offers a digital healthcare app using a mixture of artificial intelligence (AI) and video and text consultations with doctors and specialists, has raised $60 million in new funding. The company says it plans to use the new capital to continue building out its AI capabilities, including offering diagnosis by AI (rather than a more simple triage functionality), which is pegged to roll out later this year. Babylon's previous investors also include Hoxton Ventures, Richard Reed, Adam Balon and Jon Wright (co-founders of Innocent Drinks), and, perhaps most notably, Demis Hassabis and Mustafa Suleyman, founders of Deepmind, the AI group bought by Google for $500m, who are also advising the digital health startup. Earlier this year, Babylon starting working with a number of health authorities in London to trial its AI-powered chatbot'triage' service as an alternative to the NHS's 111 telephone helpline that patients call to get healthcare advice and be directed to local and out-of-hours medical services. It built on top of the same AI-enabled symptom checker feature in the main Babylon app that was released last July and which the company says has provided medical advice to over 250,000 people to date.

The paper submissions for ICLR 2017 in Toulon France deadline has arrived and instead of a trickle of new knowledge about Deep Learning we get a massive deluge. This is a gold mine of research that's hot off the presses. Many papers are incremental improvements of algorithms of the state of the art. I had hoped to find more fundamental theoretical and experimental results of the nature of Deep Learning, unfortunately there were just a few. There was however 2 developments that were mind boggling and one paper that is something I've been suspecting for a while now and has finally been confirm to shocking results.

Artificial intelligence models may be the new tool to help screen and evaluate efforts in tuberculosis-prevalent areas that often are plagued by limited access to radiologists. In TB-prone areas, there is a lack of trained radiologists qualified to screen and diagnose TB, which can be done using chest imaging techniques. However, the researchers used deep learning, a type of artificial intelligence that allows computers to complete tasks based on existing relationships of data. They modeled a deep convolutional neural network (DCNN) after brain structure to employ multiple hidden layers and patterns to classify images. "There is a tremendous interest in artificial intelligence, both inside and outside the field of medicine," Dr. Paras Lakhani, study co-author and assistant professor of Radiology at Thomas Jefferson University Hospital (TJUH) in Philadelphia, said in a statement.

This can be seen by the fact that a single repeated word commonly triggers an endless repetitive cycle. For example, a single substitution error Germany beat Germany leads to the catastrophic Germany beat Germany beat Germany beat…, and not the less-wrong Germany beat Germany 2-0. Our solution for Problem 1 (inaccurate copying) is the pointer-generator network. This is a hybrid network that can choose to copy words from the source via pointing, while retaining the ability to generate words from the fixed vocabulary. Let's step through the diagram! This diagram shows the third step of the decoder, when we have so far generated the partial summary Germany beat. As before, we calculate an attention distribution and a vocabulary distribution. However, we also calculate the generation probability, which is a scalar value between 0 and 1. This represents the probability of generating a word from the vocabulary, versus copying a word from the source.

It turns out that in addition to making spooky trailers and writing quirky scripts, artificial intelligence is also pretty damn good at making awesome movie recommendations. Available for free, Movix is a web-based recommendation service that uses deep learning algorithms to adapt to your capricious film preferences and hit you with the most suitable movie suggestions in real-time. Gary Vaynerchuk was so impressed with TNW Conference 2016 he paused mid-talk to applaud us. What is particularly nifty about the AI-powered tool is that it requires no registration, allowing you to get started in a matter of seconds. All you need to do is feed the algorithm with your favorite titles and it will immediately begin spitballing tons of movie suggestions it deems you're likely to appreciate.

We recently collaborated with InSoundz, an audio-tracking startup, to build an object detection system using Microsoft's open source deep learning framework, Computational Network Toolkit (CNTK). InSoundz captures and models 3D audio of live sports events to enhance live video feeds of these events for fans. In order to enable automatic discovery of interesting scenarios that would be relevant to their solution, InSoundz wanted to integrate object detection capabilities into their system. Any solution needed to be as flexible as possible, and also had to support adding new object types and creating detectors for various data types with ease. Since the object detection component evaluates images from a live camera feed, the detection also had to be fast, with near real-time performance. Often when people talk about "object detection," they actually mean a combination of object detection (e.g.

The best performing artificial intelligence model was a combination of the AlexNet and GoogLeNet, with a net accuracy of 96 percent. "The relatively high accuracy of the deep learning models is exciting. The applicability for TB is important because it's a condition for which we have treatment options. It's a problem that can be solved," Dr. Lakhani shared. The two DCNN models had disagreement in 13 of the 150 test cases. For these 13 cases, the scientists evaluated a workflow where an expert radiologist was able to interpret the images, accurately diagnosing 100 percent of the cases.