You can save a massive 91% off the Deep Learning and Artificial Intelligence Introductory Bundle in the Geeky Gadgets Deals store. The Deep Learning and Artificial Intelligence Introductory Bundle normally costs $480 and we have it available for $39. Deep Learning is a set of powerful algorithms that are the force behind self-driving cars, image searching, voice recognition, and many, many more applications we consider decidedly "futuristic." One of the central foundations of deep learning is linear regression; using probability theory to gain deeper insight into the "line of best fit." This is the first step to building machines that, in effect, act like neurons in a neural network as they learn while they're fed more information.

It has been rewarding and energizing to see all of the enthusiastic support about the Partnership on AI to Benefit People and Society (Partnership on AI) following our announcement in September. In the months since then, we've been working with colleagues and partners from a range of disciplines to build out a robust multi-stakeholder organization and to formulate directions for forthcoming research programs and activities. Today we have some important updates to share. Apple has joined the Partnership on AI as a founding member. The company has been involved and collaborating with the Partnership since before it was first announced and is thrilled to formalize its membership alongside Amazon, Facebook, Google/DeepMind, IBM, and Microsoft.

There's been a lot of hand-wringing about artificial intelligence and robots taking away jobs--by one recent estimate, AI could replace up to six percent of jobs in the U.S. by 2021. While most of those will be in customer service and transportation, a recent study suggests that at least one job requiring highly skilled labor could also be getting some help from AI: dermatologist. Susan Scutti at CNN reports that researchers at Stanford used a deep learning algorithm developed by Google to diagnose skin cancer. The team taught the algorithm to sort images and recognize patterns by feeding it images of everyday objects over the course of a week. "We taught it with cats and dogs and tables and chairs and all sorts of normal everyday objects," Andre Esteva, lead author on the article published this week in the journal Nature, tells Scutti. "We used a massive data set of well over a million images."

As many jobs are disappearing to automation, the latest profession to also start seeing the future may be dermatology. Stanford University researchers have developed a deep convolutional neural network, an artificial intelligence technique for building a knowledge set, to learn how to spot suspect cancer lesions. Today this process is manual and prone to errors of subjectivity. Dermatologists simply look through a dermatoscope and judge based on their education and experience. The Stanford system was given 130,000 images of skin lesions simply labeled with previously established diagnoses that included more than 2,000 diseases.

"Nanodegree" is a registered trademark of Udacity. Udacity is not an accredited university and we don't confer traditional degrees. Udacity Nanodegree programs represent collaborations with our industry partners who help us develop our content and who hire many of our program graduates.

Deep neural networks have evolved to be the state-of-the-art technique for machine learning tasks ranging from computer vision and speech recognition to natural language processing. However, deep learning algorithms are both computationally intensive and memory intensive, making them difficult to deploy on embedded systems with limited hardware resources. To address this limitation, deep compression significantly reduces the computation and storage required by neural networks. For example, for a convolutional neural network with fully connected layers, such as Alexnet and VGGnet, it can reduce the model size by 35x-49x. Even for fully convolutional neural networks such as GoogleNet and SqueezeNet, deep compression can still reduce the model size by 10x.

Businesses that adopt artificial intelligence technology to help with jobs like automating call center activity must also consider giving employees education and training so that those who are displaced by innovation can still work. In short, companies should realize that innovation can cause human pain and that they should do something to minimize it. Accenture joins the countless other analysts, technologists, and researchers who claim that the rise of artificial intelligence technologies like deep learning is ushering a new age. Deep learning, when done right, can help developers build software that can sift through mountains of data, recognize patterns, and take action. Companies like Amazon (amzn) and Google (goog) are using AI to improve their digital assistants, those voice operated helpers on smartphones and home automation hubs, said Accenture chief technology officer Paul Daugherty during a Wednesday media event.

IBM today took the wraps off a new release of PowerAI, the prepackaged bundle of deep learning frameworks that debuted last fall. With the addition of Google's TensorFlow framework, the company says its AI business model is starting to resemble the Linux distributor Red Hat. "In a sense, PowerAI makes IBM the Red Hat of deep learning," Sumit Gupta, the vice president of IBM's High Performance Computing & Data Analytics business, told Datanami. In the same way, instead of going to TensorFlow or Caffe or other websites [for deep learning frameworks], they want an enterprise-level distribution." The first release of PowerAI included "optimized" versions of Caffe-bvlc, Caffe-ibm, Caffe-nv, DIGITS, Torch, and Theano. With this release, the software includes TensorFlow 0.12, as well as Chainer, a deep learning framework that's very popular in Japan. Gupta says PowerAI--which is free and distributed as a binary for Ubuntu Linux (sorry, Red Hat)--is resonating with IBM clients, particularly those who have bought the "Minsky" Power8 servers to run deep learning and machine learning workloads on. "Enterprise customers prefer not to go to an open source website and download software and build it.

To complement existing offerings like Sparkling Water and Steam, H2O.ai is releasing Deep Water, a new tool to help businesses make deep learning a part of everyday operations. Deep Water will open up new possibilities for the TensorFlow, MXNet and Caffe communities to engage with H2O.ai. This also means that the GPU is set to become a greater part of business operations for the entire Fortune 500, not just tech companies. SriSatish Ambati, CEO of H2O.ai, says his company has found a sweet spot with predictive analytics. Ambati gave me the example of an insurance provider using H2O to analyze images of roofs and provide insights for preventative maintenance.

Recurrent neural networks (RNNs) have drawn interest from machine learning researchers because of their effectiveness at preserving past inputs for time-varying data processing tasks. To understand the success and limitations of RNNs, it is critical that we advance our analysis of their fundamental memory properties. We focus on echo state networks (ESNs), which are RNNs with simple memoryless nodes and random connectivity. In most existing analyses, the short-term memory (STM) capacity results conclude that the ESN network size must scale linearly with the input size for unstructured inputs. The main contribution of this paper is to provide general results characterizing the STM capacity for linear ESNs with multidimensional input streams when the inputs have common low-dimensional structure: sparsity in a basis or significant statistical dependence between inputs. In both cases, we show that the number of nodes in the network must scale linearly with the information rate and poly-logarithmically with the ambient input dimension. The analysis relies on advanced applications of random matrix theory and results in explicit non-asymptotic bounds on the recovery error. Taken together, this analysis provides a significant step forward in our understanding of the STM properties in RNNs.