We revisit the bias-variance tradeoff for neural networks in light of modern empirical findings. The traditional bias-variance tradeoff in machine learning suggests that as model complexity grows, variance increases. Classical bounds in statistical learning theory point to the number of parameters in a model as a measure of model complexity, which means the tradeoff would indicate that variance increases with the size of neural networks. However, we empirically find that variance due to training set sampling is roughly constant (with both width and depth) in practice. Variance caused by the non-convexity of the loss landscape is different. We find that it decreases with width and increases with depth, in our setting. We provide theoretical analysis, in a simplified setting inspired by linear models, that is consistent with our empirical findings for width. We view bias-variance as a useful lens to study generalization through and encourage further theoretical explanation from this perspective. The traditional view in machine learning is that increasingly complex models achieve lower bias at the expense of higher variance. This balance between underfitting (high bias) and overfitting (high variance) is commonly known as the bias-variance tradeoff (Figure 1). In their landmark work that initially highlighted this bias-variance dilemma in machine learning, Geman et al. (1992) suggest that larger neural networks suffer from higher variance.

Spectral features of the empirical moment matrix constitute a resourceful tool for unveiling properties of a cloud of points, among which, density, support and latent structures. It is already well known that the empirical moment matrix encodes a great deal of subtle attributes of the underlying measure. Starting from this object as base of observations we combine ideas from statistics, real algebraic geometry, orthogonal polynomials and approximation theory for opening new insights relevant for Machine Learning (ML) problems with data supported on singular sets. Refined concepts and results from real algebraic geometry and approximation theory are empowering a simple tool (the empirical moment matrix) for the task of solving non-trivial questions in data analysis. We provide (1) theoretical support, (2) numerical experiments and, (3) connections to real world data as a validation of the stamina of the empirical moment matrix approach.

Meaning may arise from an element's role or interactions within a larger system. For example, hitting nails is more central to people's concept of a hammer than its particular material composition or other intrinsic features. Likewise, the importance of a web page may result from its links with other pages rather than solely from its content. One example of meaning arising from extrinsic relationships are approaches that extract the meaning of word concepts from co-occurrence patterns in large, text corpora. The success of these methods suggest that human activity patterns may reveal conceptual organization. However, texts do not directly reflect human activity, but instead serve a communicative function and are usually highly curated or edited to suit an audience. Here, we apply methods devised for text to a data source that directly reflects thousands of individuals' activity patterns, namely supermarket purchases. Using product co-occurrence data from nearly 1.3m shopping baskets, we trained a topic model to learn 25 high-level concepts (or "topics"). These topics were found to be comprehensible and coherent by both retail experts and consumers. Topics ranged from specific (e.g., ingredients for a stir-fry) to general (e.g., cooking from scratch). Topics tended to be goal-directed and situational, consistent with the notion that human conceptual knowledge is tailored to support action. Individual differences in the topics sampled predicted basic demographic characteristics. These results suggest that human activity patterns reveal conceptual organization and may give rise to it.

Transportation and traffic are currently undergoing a rapid increase in terms of both scale and complexity. At the same time, an increasing share of traffic participants are being transformed into agents driven or supported by artificial intelligence resulting in mixed-intelligence traffic. This work explores the implications of distributed decision-making in mixed-intelligence traffic. The investigations are carried out on the basis of an online-simulated highway scenario, namely the MIT \emph{DeepTraffic} simulation. In the first step traffic agents are trained by means of a deep reinforcement learning approach, being deployed inside an elitist evolutionary algorithm for hyperparameter search. The resulting architectures and training parameters are then utilized in order to either train a single autonomous traffic agent and transfer the learned weights onto a multi-agent scenario or else to conduct multi-agent learning directly. Both learning strategies are evaluated on different ratios of mixed-intelligence traffic. The strategies are assessed according to the average speed of all agents driven by artificial intelligence. Traffic patterns that provoke a reduction in traffic flow are analyzed with respect to the different strategies.

The increasing presence of robots in industries has not gone unnoticed. Large industrial players have incorporated them into their production lines, but smaller companies hesitate due to high initial costs and the lack of programming expertise. In this work we introduce a framework that combines two disciplines, Programming by Demonstration and Automated Planning, to allow users without any programming knowledge to program a robot. The user teaches the robot atomic actions together with their semantic meaning and represents them in terms of preconditions and effects. Using these atomic actions the robot can generate action sequences autonomously to reach any goal given by the user. We evaluated the usability of our framework in terms of user experiments with a Baxter Research Robot and showed that it is well-adapted to users without any programming experience.

Amazon has said the UK will be "taking a leading role in global innovation" as it announced plans to hire 1,000 more technology, research and other skilled workers by next year. The US online retailer is to open its first office in Manchester, with room for 600 new jobs in the Hanover Building in the city's Northern Quarter – once the headquarters of the Co-operative Group. Doug Gurr, the UK manager for Amazon, said the UK was "taking a leading role in our global innovation". "These are Silicon Valley jobs in Britain, and further cement our long-term commitment to the UK," he said. Amazon said the new Manchester team would work on research and development, including software development and machine learning.

Amazon has revealed plans to create more than 1,000 jobs in the UK in Manchester, Edinburgh and Cambridge. At least 600 "highly skilled" roles will be added in Manchester working on software, machine learning and AWS, its cloud computing business. The company will also create 250 and 180 jobs at its development centres in Edinburgh and Cambridge respectively. Doug Gurr, Amazon's UK country manager, described the new roles as "Silicon Valley jobs in Britain". Liam Fox, the International Trade Secretary, said the new positions were an "enormous vote of confidence in the UK".

Retail to cloud-computing giant Amazon plans to hire over 1,000 new staff across three sites in the UK, and will open a new office in Manchester next year. "These are Silicon Valley jobs in Britain, and further cement our long-term commitment to the UK." said Doug Gurr, Amazon's UK country manager. A new corporate office in Manchester, due to open next year, will be located in the Hanover Building in the Northern Quarter. The company said the six-storey, 90,000 square-foot site will house at least 600 new staff working on software development, machine learning and R&D. Amazon said it will also expand its development centre in Edinburgh, adding 250 new staff where it already has hundreds of software engineers, machine learning scientists and user experience designers.

Looking to build up its AI portfolio, Samsung Electronics said it is acquiring Spain's Zhilabs as it charts its transition to 5G wireless networks supporting Internet of Things devices with the Spanish company's AI-based network and service analytics. Terms of the acquisition were not disclosed. The South Korean electronics giant did say it would retain Zhilabs' executive team and that the wholly-owned unit would operate independently. The deal announced on Wednesday (Oct. The deployment of "5G will enable unprecedented services attributed to the generation of exponential data traffic, for which automated and intelligent network analytics tools are vital," said Youngky Kim, president of Samsung Electronics' networking business.

Israel is becoming a world leader in medical A.I. with dozens of new health care start-ups in a country that has a population just shy of New Jersey's. The government announced a five-year program with a budget of $280 million to digitize patient data and use A.I. to gather important insights, with hopes of turning the homegrown expertise into consumer products that could make Israel an industry leader. India released its A.I. strategy only this summer, but it contains a big idea that could catch them up: become the "garage" that develops A.I. that creates economic growth and social development for themselves and the rest of the developing world. The plan, which they are calling #AIforAll, will focus on projects around health care, agriculture, education, smart cities and infrastructure, and smart mobility and transportation. The French government released a 150-page document earlier this year that spells out its A.I. efforts around the health, environment, transportation and security sectors, and is putting $2 billion into funding projects around those areas.