Internet trolls, by definition, are disruptive, combative and often unpleasant with their offensive or provocative online posts designed to disturb and upset. The common assumption is that people who troll are different from the rest of us, allowing us to dismiss them and their behavior. But research from Stanford University and Cornell University, published as part of the upcoming 2017 Conference on Computer-Supported Cooperative Work and Social Computing (CSCW 2017), suggests otherwise. The research offers evidence that, under the right circumstances, anyone can become a troll. "We wanted to understand why trolling is so prevalent today," said Justin Cheng, a computer science researcher at Stanford and lead author of the paper.

Intel CEO Brian Krzanich promised $7 billion today to resume construction of a chip factory near Phoenix that could one day employ 3,000 people. But he didn't make the announcement at the job site or on stage during a Silicon Valley keynote. Instead, Krzanich stood in the Oval Office holding a sheet of microchips next to President Trump. The photo-op played into Trump's #AmericaFirst promise of more US manufacturing jobs, and the president didn't waste time exploiting the PR moment. But everything was not as it looked.

A single conversation can be interpreted in a variety of ways. A new AI system emulates the human ability to identify emotions and the tone of speech to establish whether conversations are sad, happy or neutral. This brings scientists closer to developing a potential solution. The AI system was developed by a team of researchers at the Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Institute of Medical Engineering and Sciences (IMES) at Massachusetts Institute of Technology in Boston. In a news release, published Feb.1, the researchers explain the main functions of the system, which takes the form of an application that can be loaded into a smartwatch or wearable device.

The seminal paper of Caponnetto and de Vito (2007) provides minimax-optimal rates for kernel ridge regression in a very general setting. Its proof, however, contains an error in its bound on the effective dimensionality. In this note, we explain the mistake, provide a correct bound, and show that the main theorem remains true. The mistake lies in Proposition 3's bound on the effective dimensionalityN (λ), particularly its dependence on the parameters of the family of distributionsb and β . We discuss the mistake and provide a correct bound in Section 1.

We study an online multi-task learning setting, in which instances of related tasks arrive sequentially, and are handled by task-specific online learners. We consider an algorithmic framework to model the relationship of these tasks via a set of convex constraints. To exploit this relationship, we design a novel algorithm -- COOL -- for coordinating the individual online learners: Our key idea is to coordinate their parameters via weighted projections onto a convex set. By adjusting the rate and accuracy of the projection, the COOL algorithm allows for a trade-off between the benefit of coordination and the required computation/communication. We derive regret bounds for our approach and analyze how they are influenced by these trade-off factors. We apply our results on the application of learning users' preferences on the Airbnb marketplace with the goal of incentivizing users to explore under-reviewed apartments.

Estimation of functions of $ d $ variables is considered using ridge combinations of the form $ \textstyle\sum_{k=1}^m c_{1,k} \phi(\textstyle\sum_{j=1}^d c_{0,j,k}x_j-b_k) $ where the activation function $ \phi $ is a function with bounded value and derivative. These include single-hidden layer neural networks, polynomials, and sinusoidal models. From a sample of size $ n $ of possibly noisy values at random sites $ X \in B = [-1,1]^d $, the minimax mean square error is examined for functions in the closure of the $ \ell_1 $ hull of ridge functions with activation $ \phi $. It is shown to be of order $ d/n $ to a fractional power (when $ d $ is of smaller order than $ n $), and to be of order $ (\log d)/n $ to a fractional power (when $ d $ is of larger order than $ n $). Dependence on constraints $ v_0 $ and $ v_1 $ on the $ \ell_1 $ norms of inner parameter $ c_0 $ and outer parameter $ c_1 $, respectively, is also examined. Also, lower and upper bounds on the fractional power are given. The heart of the analysis is development of information-theoretic packing numbers for these classes of functions.

We introduce Network Maximal Correlation (NMC) as a multivariate measure of nonlinear association among random variables. NMC is defined via an optimization that infers transformations of variables by maximizing aggregate inner products between transformed variables. For finite discrete and jointly Gaussian random variables, we characterize a solution of the NMC optimization using basis expansion of functions over appropriate basis functions. For finite discrete variables, we propose an algorithm based on alternating conditional expectation to determine NMC. Moreover we propose a distributed algorithm to compute an approximation of NMC for large and dense graphs using graph partitioning. For finite discrete variables, we show that the probability of discrepancy greater than any given level between NMC and NMC computed using empirical distributions decays exponentially fast as the sample size grows. For jointly Gaussian variables, we show that under some conditions the NMC optimization is an instance of the Max-Cut problem. We then illustrate an application of NMC in inference of graphical model for bijective functions of jointly Gaussian variables. Finally, we show NMC's utility in a data application of learning nonlinear dependencies among genes in a cancer dataset.

The pathwise coordinate optimization is one of the most important computational frameworks for high dimensional convex and nonconvex sparse learning problems. It differs from the classical coordinate optimization algorithms in three salient features: {\it warm start initialization}, {\it active set updating}, and {\it strong rule for coordinate preselection}. Such a complex algorithmic structure grants superior empirical performance, but also poses significant challenge to theoretical analysis. To tackle this long lasting problem, we develop a new theory showing that these three features play pivotal roles in guaranteeing the outstanding statistical and computational performance of the pathwise coordinate optimization framework. Particularly, we analyze the existing pathwise coordinate optimization algorithms and provide new theoretical insights into them. The obtained insights further motivate the development of several modifications to improve the pathwise coordinate optimization framework, which guarantees linear convergence to a unique sparse local optimum with optimal statistical properties in parameter estimation and support recovery. This is the first result on the computational and statistical guarantees of the pathwise coordinate optimization framework in high dimensions. Thorough numerical experiments are provided to support our theory.

Sophos (LSE: SOPH), a global leader in network and endpoint security, today announced it has entered into an agreement to acquire Invincea, a visionary provider of next-generation malware protection. Invincea's endpoint security portfolio is designed to detect and prevent unknown malware and sophisticated attacks via its patented deep learning neural-network algorithms. It has been consistently ranked as among the best performing machine learning, signature-less next-generation endpoint technologies in third-party testing and rated highly both for high detection and low false-positive rates. Headquartered in Fairfax, Va., Invincea was founded by chief executive officer Anup Ghosh to address the rapidly growing zero-day security threat from nation states, cyber criminals and rogue actors. Invincea's flagship product X by Invincea uses deep learning neural networks and behavioral monitoring to detect previously unseen malware and stops attacks before damage occurs.

Researchers have found that children become better at math if their whole bodies are engaged while learning. They also found that many children improve at math if the way it's taught is individualized to each child. The research could have an impact on new teaching methods and the incorporation of physical activity during the school day. The study, conducted by researchers at the University of Copenhagen's Department of Nutrition, Exercise and Sports, investigated whether different types of math learning strategies change the way children solve math problems. The research, published in the journal Frontiers in Human Neuroscience, was conducted over a six-week period and involved testing the mathematical abilities of school children with an average age of seven years old.