The technology world is split when it comes to the future of AI. Is it going to unlock a whole new level of comfort and innovation, or is it going to destroy humanity? This next exhibit about the potential AI holds (good or bad) will literally take your breath away -- and may also help you make up your mind. And it all started off with a chess game. In a hundred game chess marathon, Google DeepMind's AlphaZero, an AI computer program, became the greatest ever chess player in the game's history by annihilating a competing AI system called Stockfish 8. Google's DeepMind merely instructed AlphaZero on the rules of the game, and asked it to learn the game of chess by playing against itself.

Aaron delivered this talk at!!Con, which is hands-down our favorite programming conference. Joel Grus's Livecoding Madness - Let's Build a Deep Learning Library was dazzling. N.K. Jemisin completed her Broken Earth trilogy with The Stone Sky, and it was the best sci-fi we read this year. While not directly related to AI, it explores consequences of a society shaped by technology inherited from the past. Autonomous by Annalee Newitz is an interesting look at bio-hacking and robot relationships. Kyle McDonald's twitter account is not just about visualization (e.g., his dispatches from NIPS were fascinating), but we particularly love it when he shares his work looking at data.

About this course: This course will teach you how to build convolutional neural networks and apply it to image data. Thanks to deep learning, computer vision is working far better than just two years ago, and this is enabling numerous exciting applications ranging from safe autonomous driving, to accurate face recognition, to automatic reading of radiology images. You will: - Understand how to build a convolutional neural network, including recent variations such as residual networks. This is the fourth course of the Deep Learning Specialization.

If you work in the area of artificial intelligence (AI) and cognitive computing, you might use buzzwords and phrases that, to others, might be perceived as confusing jargon. This article attempts to explain what these terms mean, how they relate to one other, and where they all fit along the AI and cognitive time continuum. I include a glossary of my top 20 useful AI/cognitive terms and advice on getting started on your AI/cognitive journey. Think of machine learning (ML) as a set of libraries and an execution engine for running a set of algorithms as part of a model to predict one or more outcomes. Each outcome has an associated score indicating the confidence level at which it will occur.

Worldwide, the cancer of the cervix (lower portion of the uterus) is the fourth most common cancer. It is also one of the most common causes of deaths due to cancer in women. Most of my patients that participated in my public health project had either dementia, Alzheimer's or were frail and sometimes immobile. They would forget their surroundings, spouse name and even getting a regular medical checkup was a challenge. These women, when asked to go for cervical cancer diagnosis, opted out and never showed up. Most of these tests are widely available but are uncomfortable and invasive.

I am an Artificial Intelligence Software Engineer and Data Scientist working in AI, Machine Learning, Deep Learning, and Enterprise Projects Design and Delivery. I was involved in several projects with the following technologies: C/C (11 & 14), Python, Java/J2EE, Artificial Intelligence, Machine and Deep Learning (Scikit-Learn, Tensorflow, Theano, Keras), R, Big Data, Hadoop, Spark. My main interests include Machine Learning, Deep Learning, Reinforcement Learning, Convolutional Networks and Sequence Modeling, Bio-inspired adaptive systems, self-organizing models and Neural Language Processing. I've been working in the following business contexts: Public administrations, Utilities, NATO/military organizations, Healthcare Informatics, Online advertising, and B2C services. If you're interested in my freelance services, please get in touch through the contact form.

Machine learning has already delivered remarkable results in certain niches where pattern recognition is obvious, but it's making even bigger and longer lasting impacts on businesses that demand broad insights and efficiencies in their industries. The investments of tech giants in machine learning applications are drawing a lot of attention. Google's largest collection of developers outside its US headquarters is a research group dedicated to machine learning. Microsoft open sourced CNTK, Baidu released PaddlePaddle, Amazon decided to support MXNet on AWS, and Facebook created two deep learning frameworks. The wave of machine learning applications in the consumer space will spill over into industry, which will help engineers and managers improve business operations with automated data analysis.

Multi-modal semantics, which aims to ground semantic representations in perception, has relied on feature norms or raw image data for perceptual input. In this paper we examine grounding semantic representations in raw auditory data, using standard evaluations for multi-modal semantics. After having shown the quality of such auditorily grounded representations, we show how they can be applied to tasks where auditory perception is relevant, including two unsupervised categorization experiments, and provide further analysis. We find that features transfered from deep neural networks outperform bag of audio words approaches. To our knowledge, this is the first work to construct multi-modal models from a combination of textual information and auditory information extracted from deep neural networks, and the first work to evaluate the performance of tri-modal (textual, visual and auditory) semantic models.

In many applications of classifier learning, training data suffers from label noise. Deep networks are learned using huge training data where the problem of noisy labels is particularly relevant. The current techniques proposed for learning deep networks under label noise focus on modifying the network architecture and on algorithms for estimating true labels from noisy labels. An alternate approach would be to look for loss functions that are inherently noise-tolerant. For binary classification there exist theoretical results on loss functions that are robust to label noise. In this paper, we provide some sufficient conditions on a loss function so that risk minimization under that loss function would be inherently tolerant to label noise for multiclass classification problems. These results generalize the existing results on noise-tolerant loss functions for binary classification. We study some of the widely used loss functions in deep networks and show that the loss function based on mean absolute value of error is inherently robust to label noise. Thus standard back propagation is enough to learn the true classifier even under label noise. Through experiments, we illustrate the robustness of risk minimization with such loss functions for learning neural networks.

With the availability of more computational resources and abundance of data, there has been a huge resurgence of using deep neural networks to do object recognition and classification but several machine learning models, including state-of-the-art deep neural networks, consistently misclassify adversarial examples, which are inputs formed by applying small, but intentionally engineered, worst-case perturbations to input images. These perturbations are indiscernible for humans, but they can make deep neural networks to make wrong classifications with very high confidence. The problem becomes more concerning with the advent of self-driving cars which does automatic detection and classification of road signs to do path planning, adjusting speed or driving behaviors. If the Convolutional Neural Network which detects road signs in a self-driving car is fed with adversarial inputs, even though it is obvious for a human to classify it correctly, the network may make an egregious misclassification of that road sign. This can result in self-driving cars making erroneous decisions. In this work, ways to create adversarial examples from road sign images are explored in order to use them to fool the state-of-the-art neural networks and an effort to build more robust neural networks to be resilient against these attacks is made. In Section 2, some of the previous work that has been done related to adversarial examples is addressed. Explanations of the methods that were used to craft adversarial examples and the ways used to build more robust neural networks to be resilient against adversarial samples are presented in Section 3. The dataset used and the data augmentation processes are also described in 3. Experimental results are shown in Section 4 and finally, further discussions on the weakness of this work as well as the possible future extensions of this work are discussed in Section 5. Finally, the scope of the work is concluded in Section 6.