The place where robotics, not just software, has become embedded everywhere. Everything has a smart sensor. Of course, many people refer to this trend as the Internet of Things (IoT), but we don't like that term because it doesn't tell you what those things are. These are just a few of the tens of thousands of different types of sensors in the world, all for different use cases. You can see how combining different types of data with a new kind of intelligence and hardware can give you new forms of world-changing technologies that never existed before. At the highest level, all of these things are considered robotics.

Artificial Intelligence has had a transformative impact on the modern economy. Technologies like machine learning, deep learning, and neural networks are being combined with cloud computing, open source software tools, and big data to power unprecedented breakthroughs in AI and revolutionize Finance. From compliance to traders, AI is being used across the industry to automate repetitive tasks, improve efficiency, and augment human decision making.

But as machine learning gets deeper, we are embarking on the next step towards increasingly sophisticated AI: deep learning. The sophisticated analysis of deep learning is achieved through neural networks, so called because they loosely mimic the interconnected structure of the human brain to provide a many-layered functionality. These neural networks are so sophisticated, in fact, that the path a machine takes to reach its conclusion is not yet readily understood. Deep learning uses huge, self-improving neural networks -- only possible and more widely accessible because of recent advances in computing power -- to achieve extremely complex pattern spotting like recognizing speech or images. "Deep learning is only going to be used when it really makes sense--where it can quickly find intricate, variable relationships hidden in large volumes of data that we haven't been able to pull out in any other way yet," explains Mary Beth Ainsworth, global product marketing manager of artificial intelligence and text analytics at SAS. "But deep learning means a machine can look at a problem through a completely different analytic lens than its human counterpart. It could be used to tackle all sorts of issues. The potential in all the data we collect every day is yet to be realised."

These are the first part of my notes for chapter 3 of the Deep Learning book. They can also serve as a quick intro to probability. These notes cover about half of the chapter (the part on introductory probability), a followup post will cover the rest (some more advanced probability and information theory). As usual, this post is based on a Jupyter notebook that can be found here. Perhaps one way to describe probability is as similar to logic, but when uncertainty comes in.

Deep learning is a sub-category within machine learning and artificial intelligence. It is inspired by and based on the model of the human brain to create artificial neural networks for machines. Deep learning will allow machines and devices to function in some ways as humans do. Dr. Rodrigo Agundez of GoDataDriven is co-author of this article and very enthusiastic about the improvements that deep learning can offer. He's been involved in the data science and analysis field for some time, and is already working on implementing models for practical applications.

DeepMind is an artificial intelligence research company that specializes in deep learning. That's an approach, inspired by neural networks, that passes an input through multiple, sequential layers of analysis (the "deep") to come up with an output. The method seems to be working pretty well for the company; it's the one that made AlphaGo, which it claims is "arguably the strongest Go player in history." Now that DeepMind has solved Go, the company is applying DeepMind to navigation. Navigation relies on knowing where you are in space relative to your surroundings and continually updating that knowledge as you move.

Welcome to this course: TensorFlow Slim(TF-Slim) In Depth. TF-Slim is a lightweight library for defining, training and evaluating complex models in TensorFlow. It is a library that makes building, training and evaluation neural networks simple. The course helps you obtain in-depth knowledge of TensorFlow-Slim, making you the go-to person for solving artificial intelligence problems. At the end of this course, you will have mastered the offerings of TensorFlow, and gained the skills you need to build smarter, faster, and efficient machine learning and deep learning systems.

We have all been there. You have a stellar concept that can be implemented using a machine learning model. Chances are, you find a dataset that has around a few hundred images. You recall that most popular datasets have images in the order of tens of thousands (or more). You also recall someone mentioning having a large dataset is crucial for good performance.

Now that computer scientists at Lawrence Berkeley National Laboratory's National Energy Research Scientific Computing Center (NERSC) have demonstrated 15 petaflops deep-learning training performance on the Cray Cori supercomputer, the NERSC staff is working to address the data management issues that arise when running production deep-learning codes at such scale. The existing deep learning tools were not designed to efficiently ingest or manage the terabyte- to petabyte-sized deep-learning training sets that scientists can now use on this leadership class supercomputer. "Enabling the NERSC user community to perform deep learning at scale on Cori," Quincey Koziol (Staff, Berkeley Lab) observes, "means scientists can use deep learning as part of their leading-edge scientific efforts." Thus NERSC staff are working to break new ground in adapting existing deep-learning frameworks to run efficiently at scale on thousands of nodes while giving researchers the ability to create and manage training sets containing tens to hundreds of terabytes of data in a portable fashion. For these datasets, it is imperative that they are formatted so Cori can ingest them efficiently at runtime.

A month after Google announced breakthroughs in Text-to-Speech generation technologies stemming from the Magenta project, the company followed through with a major upgrade of its Speech-to-Text API cloud service. The updated service leverages deep-learning models for speech transcription that are tailored to specific use-cases: short voice commands, phone calls and video, with a default model in all other contexts. The upgraded service now handles 120 languages and variants with different model availability and feature levels. Business applications range from over-the-phone meetings, to call-centers and video transcription. Transcription accuracy is improved in the presence of multiple speakers and significant background noise.