As far as artificial intelligence goes, 2016 has been the year of deep learning. Brain-inspired neural networks have received massive amounts of investment in time, resources and funding -- and, boy, has it ever paid off! In a new piece of research -- carried out by investigators at Rutgers University, the University of North Carolina at Charlotte, Lehigh University, and the Chinese University of Hong Kong -- neural networks have been used to generate high quality images based on nothing more detailed than basic text descriptions. "Generating realistic images from text descriptions has many applications," researcher Han Zhang told Digital Trends. "Previous approaches have difficulty in generating high resolution images, and their synthesized images in many cases lack details and vivid object parts. Our StackGAN for the first time generates 256 x 256 images with photo-realistic details."

Children come to school with very different needs and abilities, and millions of students struggle with basic reading or math skills. If teachers had more time to work with their students one-on-one, they would learn exactly where each child is having trouble. Often, that's not always possible in a typical classroom setting but this is where adaptive learning can help. Based on machine learning and artificial intelligence technologies, adaptive learning software can adjust to how students are performing in real time, changing the education model by anticipating and then delivering the specific types of learning content that students need to progress. The software acts like an intelligent tutor that responds dynamically to each child's needs and abilities, supplementing the instruction that a teacher provides and giving struggling students the personalized attention they need to succeed. The New Media Consortium's 2015 K-12 Horizon Report identified adaptive learning as one of the technologies that's likely to reach a critical mass of adoption in K-12 schools within the next few years.

Artificial intelligence (AI) is a sub-division of computer science. The main goal is to enable a smart device (e.g. First mentioned back in the 50s in the paper "Computing Machinery and Intelligence", written by mathematician Alan Turing, artificial intelligence is now a very popular field, and we have advanced technology to "blame" for that. This article is about learning Artificial Intelligence and we will give you a comprehensive guide that you can use as a starting point towards learning artificial intelligence. Today's AI-based computers can beat chess champions, so it's safe to say that little by little the world is taking a turn. Some people say that artificial intelligence will save humanity; others, claim it will destroy it.

If We Want Machines to Think, We Need to Teach Them to See. Learning and computation provides machine the ability to better understand the context of images and build visual systems which truly understand intelligence. The huge amount of image and video content urges the scientific community to make sense and identify patterns amongst it to reveal details which we aren't aware of. Computer Vision generates mathematical models from images; Computer Graphics draws in images from models and lastly image processing takes image as an input and gives an image at the output. Computer Vision is an overlapping field drawing on concepts from areas such as artificial intelligence, digital image processing, machine learning, deep learning, pattern recognition, probabilistic graphical models, scientific computing and a lot of mathematics.

One crucial goal in kernel online learning is to bound the model size. Common approaches employ budget maintenance procedures to restrict the model sizes using removal, projection, or merging strategies. Although projection and merging, in the literature, are known to be the most effective strategies, they demand extensive computation whilst removal strategy fails to retain information of the removed vectors. An alternative way to address the model size problem is to apply random features to approximate the kernel function. This allows the model to be maintained directly in the random feature space, hence effectively resolve the curse of kernelization. However, this approach still suffers from a serious shortcoming as it needs to use a high dimensional random feature space to achieve a sufficiently accurate kernel approximation. Consequently, it leads to a significant increase in the computational cost. To address all of these aforementioned challenges, we present in this paper the Dual Space Gradient Descent (DualSGD), a novel framework that utilizes random features as an auxiliary space to maintain information from data points removed during budget maintenance. Consequently, our approach permits the budget to be maintained in a simple, direct and elegant way while simultaneously mitigating the impact of the dimensionality issue on learning performance. We further provide convergence analysis and extensively conduct experiments on five real-world datasets to demonstrate the predictive performance and scalability of our proposed method in comparison with the state-of-the-art baselines.

In online convex optimization it is well known that certain subclasses of objective functions are much easier than arbitrary convex functions. We are interested in designing adaptive methods that can automatically get fast rates in as many such subclasses as possible, without any manual tuning. Previous adaptive methods are able to interpolate between strongly convex and general convex functions. We present a new method, MetaGrad, that adapts to a much broader class of functions, including exp-concave and strongly convex functions, but also various types of stochastic and non-stochastic functions without any curvature. For instance, MetaGrad can achieve logarithmic regret on the unregularized hinge loss, even though it has no curvature, if the data come from a favourable probability distribution. MetaGrad's main feature is that it simultaneously considers multiple learning rates. Unlike all previous methods with provable regret guarantees, however, its learning rates are not monotonically decreasing over time and are not tuned based on a theoretically derived bound on the regret. Instead, they are weighted directly proportional to their empirical performance on the data using a tilted exponential weights master algorithm.

Matrix completion is a basic machine learning problem that has wide applications, especiallyin collaborative filtering and recommender systems. Simple non-convex optimization algorithms are popular and effective in practice. Despite recent progress in proving various non-convex algorithms converge from a good initial point, it remains unclear why random or arbitrary initialization suffices in practice. We prove that the commonly used non-convex objective function for positive semidefinite matrix completion has no spurious local minima - all local minima must also be global. Therefore, many popular optimization algorithms such as (stochastic) gradient descent can provably solve positive semidefinite matrix completion with arbitrary initialization in polynomial time. The result can be generalized to the setting when the observed entries contain noise. We believe that our main proof strategy can be useful for understanding geometric properties of other statistical problems involving partial or noisy observations.

Many practical perception systems exist within larger processes which often include interactions with users or additional components that are capable of evaluating the quality of predicted solutions. In these contexts, it is beneficial to provide these oracle mechanisms with multiple highly likely hypotheses rather than a single prediction. In this work, we pose the task of producing multiple outputs as a learning problem over an ensemble of deep networks -- introducing a novel stochastic gradient descent based approach to minimize the loss with respect to an oracle. Our method is simple to implement, agnostic to both architecture and loss function, and parameter-free. Our approach achieves lower oracle error compared to existing methods on a wide range of tasks and deep architectures. We also show qualitatively that solutions produced from our approach often provide interpretable representations of task ambiguity.

Deep neural networks continue to advance the state-of-the-art of image recognition tasks with various methods. However, applications of these methods to multimodality remain limited. We present Multimodal Residual Networks (MRN) for the multimodal residual learning of visual question-answering, which extends the idea of the deep residual learning. Unlike the deep residual learning, MRN effectively learns the joint representation from visual and language information. The main idea is to use element-wise multiplication for the joint residual mappings exploiting the residual learning of the attentional models in recent studies. Various alternative models introduced by multimodality are explored based on our study. We achieve the state-of-the-art results on the Visual QA dataset for both Open-Ended and Multiple-Choice tasks. Moreover, we introduce a novel method to visualize the attention effect of the joint representations for each learning block using back-propagation algorithm, even though the visual features are collapsed without spatial information.

We consider a crowdsourcing model in which n workers are asked to rate the quality of n items previously generated by other workers. An unknown set of $\alpha n$ workers generate reliable ratings, while the remaining workers may behave arbitrarily and possibly adversarially. The manager of the experiment can also manually evaluate the quality of a small number of items, and wishes to curate together almost all of the high-quality items with at most an fraction of low-quality items. Perhaps surprisingly, we show that this is possible with an amount of work required of the manager, and each worker, that does not scale with n: the dataset can be curated with $\tilde{O}(1/\beta\alpha\epsilon^4)$ ratings per worker, and $\tilde{O}(1/\beta\epsilon^2)$ ratings by the manager, where $\beta$ is the fraction of high-quality items. Our results extend to the more general setting of peer prediction, including peer grading in online classrooms.