A Convolutional Neural Network (CNN) is comprised of one or more convolutional layers (often with a subsampling step) and then followed by one or more fully connected layers as in a standard multilayer neural network. The architecture of a CNN is designed to take advantage of the 2D structure of an input image (or other 2D input such as a speech signal). This is achieved with local connections and tied weights followed by some form of pooling which results in translation invariant features. Another benefit of CNNs is that they are easier to train and have many fewer parameters than fully connected networks with the same number of hidden units. In this article we will discuss the architecture of a CNN and the back propagation algorithm to compute the gradient with respect to the parameters of the model in order to use gradient based optimization.

Chatbots and A.I. are some of the leading innovations that are taking the tech industry by storm. Companies are pushing to better incorporate these tools to help them streamline, organize, and better communicate with their consumers. But how can these tools help companies who are already at the forefront of their niche? Slack, one of the leading companies for real-time collaboration and digital communication, is pushing towards the chatbot revolution as well. After recently launching their own VC fund and app store, Slack is taking things to the next level with investing an unspecified amount in a wide variety of chatbot companies, from Automat to Butter and Sudo.

Earlier this month, Apple made a splash when it told the artificial intelligence research community that the secretive company would start publishing AI papers of its own. Not even a month later, it's already starting to make good on that promise. Apple has published its very first AI paper on December 22. (The paper was submitted for publication on November 15.) The paper describes a technique for how to improve the training of an algorithm's ability to recognize images using computer-generated images rather than real-world images. In machine learning research, using synthetic images (like those from a video game) to train neural networks can be more efficient than using real-world images.

A 2017 Chrysler Pacifica hybrid minivan equipped with Waymo's self-driving vehicle technology. Waymo, the company born from Alphabet's Google Self-Driving Car research project, faces mounting competition to perfect technology needed for fully autonomous vehicles. After staying low key about its progress, the latest indications from the new company are that it's far along the path to making such vehicles a reality by taking cost out of the components and boosting overall performance and reliability. John Krafcik, Waymo's chief executive officer, said at the Automobili-D conference in Detroit that the latest sensors, software, artificial intelligence and other components -- all developed and built in-house -- are being used for a fleet of 100 Chrysler Pacifica minivans, the first batch of which will begin public road tests in California and Arizona this month, he said. Keeping development and production in-house has led to major cost savings, including a 90% reduction for the laser Lidar sensor riding atop the new Pacificas.

What to watch for in distributed systems, SRE, serverless, containers and more. Forecasting trends is tricky, especially in the fast-moving world of systems operations and engineering. This year, at our Velocity Conference, we have talked about distributed systems, SRE, containerization, serverless architectures, burnout, and many other topics related to the human and technological challenges of delivering software. We think this is important enough that we re-focused the entire Velocity conference on it. Site Reliability Engineering--is it just ops?

Maybe you'd love to dig deeper and get an image recognition program running in TensorFlow or Theano? Perhaps you're a kick-ass developer or systems architect and you know computers incredibly well but there's just one little problem: I share your dirty little secret and I have some books and websites that will really help you get rolling fast. Like many folks, my love of intelligent machines didn't come from calculus class. It sprang from science fiction. I remember reading "I, Robot" one beautiful summer evening and imagining ways to trip up Asimov's Three Rules of Robotics.

Artificial Intelligence (AI) and Machine Learning (ML) are two very hot buzzwords right now, and often seem to be used interchangeably. They are not quite the same thing, but the perception that they are can sometimes lead to some confusion. So I thought it would be worth writing a piece to explain the difference. Both terms crop up very frequently when the topic is Big Data, analytics, and the broader waves of technological change which are sweeping through our world.

It was actually about when technology goes wrong: In many ways, 2016 was the year of the exploding batteries. A little over a year ago, hoverboards topped many holiday wish lists. By December 2015, they were being recalled by the thousands. According to Popular Science, "…cheaply made hoverboards have exploded and caught fire, forcing Amazon to stop selling specific models and Overstock to discontinue all sales." Rolling into 2016, major computer companies recalled batteries for fire hazards, baby monitors were pulled from shelves and major airlines diverted flights for emergency landings when tablets caught fire onboard.

Rippleshot is a next generation fraud analytics firm. They use a cloud-based technology solution that takes a big data approach by leveraging machine learning/artificial intelligence to distinguish fraudulent activity more quickly and efficiently (and pro-actively). Rippleshot's technology processes millions of payment card transactions to proactively pinpoint when and where a data breach occurred. Following detection, Rippleshot provides banks with the tools they need to update fraud detection rules in order to lower their fraud losses while avoiding unnecessary card re-issuance. Rippleshot protects consumer credit information and the integrity of the merchant payment network by proactively detecting data breaches through a cloud-based solution.

NVIDIA (NASDAQ: NVDA) is primarily known as the company that revolutionized computer gaming. The debut of the Graphics Processing Unit (GPU) in 1999 provided gamers with faster, clearer, and more lifelike images. The GPU was designed to quickly perform complex mathematical calculations that were necessary to accelerate the creation of realistic graphics. It achieved this feat by performing many functions at the same time, known as parallel computing. This resulted in faster, smoother motion in game graphics and a revolution in modern gaming.