Technology / What countries are investing in robots? A recent study by Redwood Software and the Centre for Economic and Business Research (Cebr) has focused on the impact of robotics automation on economic development across OECD countries, including the UK and the US. How does UK robotics investment compare to other countries? The report found that the UK has a much smaller robotics investment/GDP share in comparison to Japan, Germany and the US. Expressed in 2015 PPP (Purchasing Power Parity) terms, robotics investment in the US was $86billion in 2015, approximately 62 times that of the UK, recovering strongly from less than $30billion during the recession in 2009.

By the year 2021, there will be more AI-powered digital assistants installed on devices than there are people in the world, according to new research from Ovum. The install base will be higher than 7.5 billion by that time, which is greater that the planet's population as recorded by the US Census Bureau on May 1, 2017. Google Assistant will be the most installed assistant, accounting for 23.3% of the market, the report said. The next most popular assistant will be Samsung Bixby, with 14.5% market share. SEE: Why robots and AI won't replace most jobs any time soon "Ultimately, a digital assistant is just another user interface. It will only be as good as the ecosystem of devices and services that it is compatible with. Partnerships between tech giants and local service providers will therefore be key differentiators," said Ronan de Renesse, practice leader for Ovum's Consumer Technology team and author of the report, in the release.

Conversations with Siri have almost become a pastime as people find new ways to garner funny responses from the iPhone personal assistant. A new study indicates that time spent chilling with your phone, or other human-like gadgets like Amazon's Alexa, could actually hinder your IRL relationships. Those who are lonely typically spend time with real people as a way to feel better. However, researchers say that devices mimicking realistic personal responses are now taking the place of, well, their actual human counterparts. "Generally, when people feel socially excluded, they seek out other ways of compensating, like exaggerating their number of Facebook friends or engaging in prosocial behaviors to seek out interaction with other people," says Jenny Olson, study co-author and marketing professor at the University of Kansas, in a statement.

We consider abstract-argumentation-theoretic coalition formability in this work. Taking a model from political alliance among political parties, we will contemplate profitability, and then formability, of a coalition. As is commonly understood, a group forms a coalition with another group for a greater good, the goodness measured against some criteria. As is also commonly understood, however, a coalition may deliver benefits to a group X at the sacrifice of something that X was able to do before coalition formation, which X may be no longer able to do under the coalition. Use of the typical conflict-free sets of arguments is not very fitting for accommodating this aspect of coalition, which prompts us to turn to a weaker notion, conflict-eliminability, as a property that a set of arguments should primarily satisfy. We require numerical quantification of attack strengths as well as of argument strengths for its characterisation. We will first analyse semantics of profitability of a given conflict-eliminable set forming a coalition with another conflict-eliminable set, and will then provide four coalition formability semantics, each of which formalises certain utility postulate(s) taking the coalition profitability into account.

Matrices of low rank are pervasive in big data, appearing in recommender systems, movie preferences, topic models, medical records, and genomics. While there is a vast literature on how to exploit low rank structure in these datasets, there is less attention on explaining why the low rank structure appears in the first place. We explain the abundance of low rank matrices in big data by proving that certain latent variable models associated to piecewise analytic functions are of log-rank. A large matrix from such a latent variable model can be approximated, up to a small error, by a low rank matrix.

We introduce a novel framework for adversarial training where the target distribution is annealed between the uniform distribution and the data distribution. We posited a conjecture that learning under continuous annealing in the nonparametric regime is stable irrespective of the divergence measures in the objective function and proposed an algorithm, dubbed {\ss}-GAN, in corollary. In this framework, the fact that the initial support of the generative network is the whole ambient space combined with annealing are key to balancing the minimax game. In our experiments on synthetic data, MNIST, and CelebA, {\ss}-GAN with a fixed annealing schedule was stable and did not suffer from mode collapse.

We consider the problem of estimation of a low-rank matrix from a limited number of noisy rank-one projections. In particular, we propose two fast, non-convex \emph{proper} algorithms for matrix recovery and support them with rigorous theoretical analysis. We show that the proposed algorithms enjoy linear convergence and that their sample complexity is independent of the condition number of the unknown true low-rank matrix. By leveraging recent advances in low-rank matrix approximation techniques, we show that our algorithms achieve computational speed-ups over existing methods. Finally, we complement our theory with some numerical experiments.

This document is written with the intention to describe in detail a method and means by which a computer program can reason about the world and in so doing, increase its analogue to a living system. As the literature is rife and it is apparent we, as scientists and engineers, have not found the solution, this document will attempt the solution by grounding its intellectual arguments within tenets of human cognition in Western philosophy. The result will be a characteristic description of a method to describe an artificial system analogous to that performed for a human. The approach was the substance of my Master's thesis, explored more deeply during the course of my postdoc research. It focuses primarily on context awareness and choice set within a boundary of available epistemology, which serves to describe it. Expanded upon, such a description strives to discover agreement with Kant's critique of reason to understand how it could be applied to define the architecture of its design. The intention has never been to mimic human or biological systems, rather, to understand the profoundly fundamental rules, when leveraged correctly, results in an artificial consciousness as noumenon while in keeping with the perception of it as phenomenon.

In this paper we study stochastic quasi-Newton methods for nonconvex stochastic optimization, where we assume that noisy information about the gradients of the objective function is available via a stochastic first-order oracle (SFO). We propose a general framework for such methods, for which we prove almost sure convergence to stationary points and analyze its worst-case iteration complexity. When a randomly chosen iterate is returned as the output of such an algorithm, we prove that in the worst-case, the SFO-calls complexity is $O(\epsilon^{-2})$ to ensure that the expectation of the squared norm of the gradient is smaller than the given accuracy tolerance $\epsilon$. We also propose a specific algorithm, namely a stochastic damped L-BFGS (SdLBFGS) method, that falls under the proposed framework. {Moreover, we incorporate the SVRG variance reduction technique into the proposed SdLBFGS method, and analyze its SFO-calls complexity. Numerical results on a nonconvex binary classification problem using SVM, and a multiclass classification problem using neural networks are reported.

We consider the stochastic composition optimization problem proposed in \cite{wang2017stochastic}, which has applications ranging from estimation to statistical and machine learning. We propose the first ADMM-based algorithm named com-SVR-ADMM, and show that com-SVR-ADMM converges linearly for strongly convex and Lipschitz smooth objectives, and has a convergence rate of $O( \log S/S)$, which improves upon the $O(S^{-4/9})$ rate in \cite{wang2016accelerating} when the objective is convex and Lipschitz smooth. Moreover, com-SVR-ADMM possesses a rate of $O(1/\sqrt{S})$ when the objective is convex but without Lipschitz smoothness. We also conduct experiments and show that it outperforms existing algorithms.