UPDATE: Mar 20, 2016 - Added my new follow-up course on Deep Learning, which covers ways to speed up and improve vanilla backpropagation: momentum and Nesterov momentum, adaptive learning rate algorithms like AdaGrad and RMSProp, utilizing the GPU on AWS EC2, and stochastic batch gradient descent. We look at TensorFlow and Theano starting from the basics - variables, functions, expressions, and simple optimizations - from there, building a neural network seems simple! Deep learning is all the rage these days. What exactly is deep learning? Well, it all boils down to neural networks.

An unnamed regionally-managed health care provider partnered with ThingWorx's machine learning platform to detect patterns in data that would lead to better patient care and reduce costly readmissions for patients with ischemic heart disease, according to a case study provided by Thingworx. The solution predicts high-risk patients and provides caregivers insight into why flagged patients should receive extra care across their network using health care IoT. The unspecified health care network includes two major hospitals and a network of outpatient and preventative care providers. It has more than 1,000 patient beds, a home health care service, preventive medicine, rehabilitation services, a network of primary care physicians and a range of outpatient services. According to Thingworx, its client is one of the largest health providers in the country.

THE DEEP LEARNING people at Yahoo have set themselves and their supercomputers the task of looking at potentially rude pictures and giving them a'NSFW' or'OK for grandma' stamp. Teenagers do this every day as they frantically scan the bushes by the side of main roads for scraps of magazines and images that look like flesh. If machine learning has taught us one thing, it is that machines need to learn to be more like us. Boing Boing reported on a Yahoo project called open_nsfw that has a project page on GitHub. Teenagers can go back to collecting pocket monsters now, because it is low on NSFW shots, although that is the whole point of the thing.

They finally did it: They made "artificial intelligence" a buzzword. Typically, buzzwords don't come from decades-old evolving disciplines of computer science. Making "machine learning," "AI" and "neural nets" into buzzwords means that millions of developers are likely having their first experience with this stuff now. In that vein, we bring you a nice long list of machine learning, deep learning, neural network and artificial intelligence how-to's. Buzzword or not, it's fairly obvious this stuff will be a big part of enterprise software for the next few decades.

Artificial intelligence is a $15 billion dollar industry and growing. With more than 2,600 companies developing intelligent technology, the value of AI is expected to rise to more than $70 billion by 2020. And it's not just attracting the tech giants: USAA is using AI to protect its users from identity theft, and Under Armour has connected its health app, MyFitnessPal, to IBM Watson so users can get a more thorough read of their health. For designers, that represents a major business opportunity. But AI is also a challenge requiring every strength and skill they've learned and many they haven't.

AT&T Labs is one of the most vital research and product development centers in the industry. To lead this elite group requires superior technical expertise– in areas like artificial intelligence (AI), networking and data analytics – as well as top-notch management skills. Mazin Gilbert, our new vice president of Labs, embodies all those qualities. Now I've known Mazin since 2006, but he's been with AT&T for more than 20 years. Mazin has played a critical role in developing many cutting-edge digital creations we take for granted.

On-board processing will give every digital eye a very powerful brain. San Mateo-based Movidius may still be in the process of getting bought up by Intel, but the company's latest deal will put its low-power AI and computer vision platform into more than just DJI drones and Google VR headsets. The company announced today that the Movidius Myriad 2 Video Processing Unit (VPU) will soon power a new generation of Hikvision smart surveillance cameras capable of recognizing everything from suspicious packages to distracted drivers. While most deep-learning neural networks require a lot of cloud-based processing power, the same platform found in Movidius' Fathom AI-on-a-stick will allow Hikvision cameras to do more on-board processing. Hikvision's cameras have already been able to achieve around 99 percent accuracy in scenarios like identifying car models, detecting intruders, spotting suspicious baggage and even calling out drivers who don't buckle up.

In another five years, Watson will be helping a doctor diagnose a patient's symptoms and a company CEO calculate whether to buy a competitor. "The technology is not even moving fast. It's moving faster and faster every day," said John Kelly III, senior vice president of Cognitive Solutions and IBM Research. Ever since Watson, an intelligent system that uses machine learning and natural language recognition, beat Jeopardy champions in 2011, the system has been used in a variety of industries, and IBM is hoping to show how far it has come. However, Kelly wanted to focus on where Watson is going.

All the problems discussed in the previous posts can create two major cross-sectional problems: the likely event to run out of money before hitting relevant milestones toward the next investment, as well as whether pursuing specific business applications to break even instead of focusing on product development. In terms instead of classifying different companies operating in the space, there might be several different ways to think around machine intelligence startups (e.g., the classification proposed by Bloomberg Beta investor Shivon Zilis in 2015 is very accurate and useful for this purpose). The solutions usually provided might either integrate with the clients' stack (through APIs or building specifically on top of customers' platform) or otherwise full-stacks solutions. Virtual agents and chatbots cover the low-cost side of the group, while physical world systems (e.g., self-driving cars, sensors, etc.), drones, and actual robots are the capital and talent-intensive side of the coin. The results of this categorization can be summarized into the following matrix, plotting the groups with respect to short-term monetization (STM) and business defensibility.

Almost all the industries including manufacturing, healthcare, construction, online retail, etc. Machine learning technology is constantly evolving and the current trends in the field promise that every enterprise will be data driven and will have the capacity of using machine learning in the cloud to incorporate artificial intelligence apps. The three newest machine learning trends that will make this possible are Data Flywheels, The Algorithm Economy, and Cloud Hosted Intelligence. The coming age of artificial intelligence will include mining of medical records to provide better and faster health services.