In this work, we investigate the value of employing deep learning for the task of wireless signal modulation recognition. Recently in [1], a framework has been introduced by generating a dataset using GNU radio that mimics the imperfections in a real wireless channel, and uses 10 different modulation types. Further, a convolutional neural network (CNN) architecture was developed and shown to deliver performance that exceeds that of expert-based approaches. Here, we follow the framework of [1] and find deep neural network architectures that deliver higher accuracy than the state of the art. We tested the architecture of [1] and found it to achieve an accuracy of approximately 75% of correctly recognizing the modulation type. We first tune the CNN architecture of [1] and find a design with four convolutional layers and two dense layers that gives an accuracy of approximately 83.8% at high SNR. We then develop architectures based on the recently introduced ideas of Residual Networks (ResNet [2]) and Densely Connected Networks (DenseNet [3]) to achieve high SNR accuracies of approximately 83.5% and 86.6%, respectively. Finally, we introduce a Convolutional Long Short-term Deep Neural Network (CLDNN [4]) to achieve an accuracy of approximately 88.5% at high SNR.

The Fisher information metric is an important foundation of information geometry, wherein it allows us to approximate the local geometry of a probability distribution. Recurrent neural networks such as the Sequence-to-Sequence (Seq2Seq) networks that have lately been used to yield state-of-the-art performance on speech translation or image captioning have so far ignored the geometry of the latent embedding, that they iteratively learn. We propose the information geometric Seq2Seq (GeoSeq2Seq) network which abridges the gap between deep recurrent neural networks and information geometry. Specifically, the latent embedding offered by a recurrent network is encoded as a Fisher kernel of a parametric Gaussian Mixture Model, a formalism common in computer vision. We utilise such a network to predict the shortest routes between two nodes of a graph by learning the adjacency matrix using the GeoSeq2Seq formalism; our results show that for such a problem the probabilistic representation of the latent embedding supersedes the non-probabilistic embedding by 10-15\%.

We study the necessary and sufficient complexity of ReLU neural networks-in terms of depth and number of weights-required for approximating classifier functions in an $L^2$-sense. As a model, we consider the set $\mathcal{E}^\beta (\mathbb{R}^d)$ of possibly discontinuous piecewise $C^\beta$ functions $f : [-1/2, 1/2]^d \to \mathbb{R}$, where the different 'smooth regions' of $f$ are separated by $C^\beta$ hypersurfaces. For given dimension $d \geq 2$, regularity $\beta > 0$, and accuracy $\varepsilon > 0$, we construct ReLU neural networks that approximate functions from $\mathcal{E}^\beta(\mathbb{R}^d)$ up to an $L^2$ error of $\varepsilon$. The constructed networks have a fixed number of layers, depending only on $d$ and $\beta$ and they have $O(\varepsilon^{-2(d-1)/\beta})$ many nonzero weights, which we prove to be optimal. In addition to the optimality in terms of the number of weights, we show that in order to achieve this optimal approximation rate, one needs ReLU networks of a certain minimal depth. Precisely, for piecewise $C^\beta(\mathbb{R}^d)$ functions, this minimal depth is given-up to a multiplicative constant-by $\beta/d$. Up to a log factor, our constructed networks match this bound. This partly explains the benefits of depth for ReLU networks by showing that deep networks are necessary to achieve efficient approximation of (piecewise) smooth functions. Finally, we analyze approximation in high-dimensional spaces where the function $f$ to be approximated can be factorized into a smooth dimension reducing feature map $\tau$ and classifier function $g$-defined on a low-dimensional feature space-as $f = g \circ \tau$. We show that in this case the approximation rate depends only on the dimension of the feature space and not the input dimension.

At some fundamental level, no one understands machine learning. It isn't a matter of things being too complicated. Almost everything we do is fundamentally very simple. Unfortunately, an innate human handicap interferes with us understanding these simple things. Humans evolved to reason fluidly about two and three dimensions. With some effort, we may think in four dimensions. Machine learning often demands we work with thousands of dimensions – or tens of thousands, or millions! Even very simple things become hard to understand when you do them in very high numbers of dimensions. Reasoning directly about these high dimensional spaces is just short of hopeless.

Over the next five years, enterprises will move closer to adopting immersive technologies such as augmented reality (AR), virtual reality (VR) and mixed reality (MR). These technologies will in turn force vendors to figure out how to get more artificial intelligence (AI) functionality out of the cloud and into the edge. When discussing immersive technologies, a fundamental point emerges: Both immersive technologies and AI are actually a collection of subset technologies. Tuong Nguyen, principal research analyst at Gartner, says businesses need to consider both immersive technologies and AI to be mutually beneficial. As AI improves, so do immersive technologies, and vice versa.

These are our first featured articles and resources posted in 2018. It covers IoT, data science, deep learning, data security, optimization, deep learning, SQL, Tensorflow and more.

Tensorflow, developed by Google, has become the most popular framework for deep learning, and now operates on a variety of devices such as multicore CPUs, general purpose GPUs, mobile devices, and custom ASICs. In this webinar co-hosted by Intel, you will get a general introduction to working with Tensorflow and its surrounding ecosystem, general problem classes, where you can get big acceleration, and why run on a CPU.

Eun Chang is a lifelong learner, an avid consumer of new technologies, and a data scientist at Microsoft specializing in machine learning. In her career at Microsoft, Eun has driven a number of projects focusing on Windows post sales monetization and enterprise device usage. Her research uses a broad spectrum of techniques ranging from statistical inference to deep learning. Eun is thrilled when a seemingly simple problem results in a novel, mathematically intricate solution. Her current work involves incorporating neural networks into areas such as feedback classification and community detection.

On November 2016, I predicted that: "we are just at the beginning of the golden age of analytics, in which the value and contributions of AI [artificial intelligence], machine learning [ML] and deep learning will only grow as we accept and incorporate these tools into our businesses." As it happens, AI, ML and deep learning didn't just grow up during 2017. In 2018, development and use of these technologies will continue to expand and flourish, especially in the finance sector. Here are some of the top things I think we'll see among the many flavors of AI in 2018. In 2018, Defensive AI will be front and center.

I consider myself strong in algorithms, data structures and programming. From last few years, I was interested to be an expert in Deep Learning Technologies. My initial understanding was that to be a good consultant in Deep Learning, I need to learn too many things to make the application work, write monstrous code using a lot of APIs, understand lot of mathematics (calculus, algebra and probability) to grasp the concepts. But my passion to master this new technology made me jump into it a year back. Then there was no looking back.