Autonomous agents must often deal with conflicting requirements, such as completing tasks using the least amount of time/energy, learning multiple tasks, or dealing with multiple opponents. In the context of reinforcement learning~(RL), these problems are addressed by (i)~designing a reward function that simultaneously describes all requirements or (ii)~combining modular value functions that encode them individually. Though effective, these methods have critical downsides. Designing good reward functions that balance different objectives is challenging, especially as the number of objectives grows. Moreover, implicit interference between goals may lead to performance plateaus as they compete for resources, particularly when training on-policy. Similarly, selecting parameters to combine value functions is at least as hard as designing an all-encompassing reward, given that the effect of their values on the overall policy is not straightforward. The later is generally addressed by formulating the conflicting requirements as a constrained RL problem and solved using Primal-Dual methods. These algorithms are in general not guaranteed to converge to the optimal solution since the problem is not convex. This work provides theoretical support to these approaches by establishing that despite its non-convexity, this problem has zero duality gap, i.e., it can be solved exactly in the dual domain, where it becomes convex. Finally, we show this result basically holds if the policy is described by a good parametrization~(e.g., neural networks) and we connect this result with primal-dual algorithms present in the literature and we establish the convergence to the optimal solution.

This paper investigates statistical models for road traffic modeling. The proposed methodology considers road traffic as a (i) highdimensional time-series for which (ii) regeneration occurs at the end of each day. Since (ii), prediction is based on a daily modeling of the road traffic using a vector autoregressive model that combines linearly the past observations of the day. Considering (i), the learning algorithm follows from an l1-penalization of the regression coefficients. Excess risk bounds are established under the high-dimensional framework in which the number of road sections goes to infinity with the number of observed days. Considering floating car data observed in an urban area, the approach is compared to state-of-the-art methods including neural networks. In addition of being very competitive in terms of prediction, it enables to identify the most determinant sections of the road network.

When Warner Brothers acquired movie discovery site, Flixster, and its reviews site, Rotten Tomatoes, in 2011, its President and COO, Steve Polsky, set out to explore how scoring consumer behavior data from mobile activity could be leveraged to solve a growing global financial crisis, one where two-thirds of the world's population have been excluded from transacting in the digital economy because they lack formal financial identity. To address this problem, Polsky founded Juvo to serve as a data-driven micro-financing provider that uses machine learning to extend emergency airtime loans to prepaid phone users when they run low on their balance. As customers make timely payments, they unlock access to progressively larger loans and build credit that they can use to qualify for a credit card and other financial instruments. Inspired by the World Bank challenge to achieve Universal Financial Access for a billion people by 2020, this San Francisco-based Series B startup has raised $54 million with rounds led by NEA and Wing, and strategic partnerships with Samsung NEXT, Nokia, Cable & Wireless and other leading mobile operators around the world. At the heart of their offerings is the data set they're mining which boasts over 5.6 billion data points being constructed real time for over 250 million people daily across 26 countries and 4 continents.

Agriculture is set to become'smarter' with digitisation and new technologies such as facial recognition for animals promising to provide the sector with greater control over its processes. Dr. Venkat Maroju, Chief Executive Officer of'SourceTrace' a provider of software solutions to the agriculture and allied sectors, said that digital technologies have an "enormous potential" to impact agriculture in several aspects by enabling farmers, on the one hand, to produce more while at the same time reducing the environmental impact of agricultural production. "Digital technology comes with several solutions to make this happen from farm management and traceability to certification and market linkage," he said. In September, itelligence AG the SAP software and technologies services company in partnership with the German Technical University OWL and HARTING Foundation & Co. KG a provider of industrial interconnection technology announced the launch of HARTING MICA . This is a system designed to enable the most efficient method of farming a given acreage of wheat using data collected via sensors from the soil, agricultural machinery and satellite images.

Gartner, Inc. today revealed its top strategic predictions for 2020 and beyond. Gartner's top predictions examine how the human condition is being challenged as technology creates varied and ever-changing expectations of humans. "Technology is changing the notion of what it means to be human," said Daryl Plummer, distinguished vice president and Gartner Fellow. "As workers and citizens see technology as an enhancement of their abilities, the human condition changes as well. CIOs in end-user organizations must understand the effects of the change and reset expectations for what technology means."

A recent study conducted by Oracle and research firm Future Workplace found that 64% of people would trust a robot more than their manager. The study included 8,370 employees, managers and HR leaders across 10 countries. Its aim was to see how AI has changed relationships between people and technology at work. It did have some surprising results when comparing human supervisors to potential robot overlords. According to the study, 64 % of people would trust a robot over their manager.

Google has updated its search algorithms to tap into an AI language model that is better at understanding netizens' queries than previous systems. Pandu Nayak, a Google fellow and vice president of search, announced this month that the Chocolate Factory has rolled out BERT, short for Bidirectional Encoder Representations from Transformers, for its most fundamental product: Google Search. To pull all of this off, researchers at Google AI built a neural network known as a transformer. The architecture is suited to deal with sequences in data, making them ideal for dealing with language. To understand a sentence, you must look at all the words in it in a specific order.

In a brightly coloured shipping container in east London, Rubens Filho is asking me to pick a card. "Any card," he says, fanning the pack out face down. "And don't worry, you can show me. I pull out the seven of spades, and show it to him; he gets me to sign my name on it with a marker pen. Then he slides it back into the middle of the pack, puts the cards back into their box and puts the box on the table in front of us. "Now," he says with a grin, "the magic begins." Filho is 51, tall, handsome and infectiously enthusiastic about the power of magic tricks and illusions. Born in Brazil, he's been a keen magician since adolescence. He came to Britain in 2012 to work in advertising, before, in 2015, setting up Abracademy, a startup dedicated to bringing magic – and in particular the skills needed to perform it – to the rest of us. "I think magic has a such a positive twist," he says. "It brings this soft approach that's hard to explain, this role of creating something beautiful." But he is also fascinated by the relationship between magic and neuroscience and psychology, and set up Abracademy Labs, an offshoot of Abracademy, to explore this connection. "Magic has lived in the'glitches' of the brain for a long time," he says. "How you see things, how you form beliefs, how you experience wonder.

The reports provide market insights into demand drivers, regional outlook, and competitive analysis of the Cognitive Computing market for the Cognitive Computing forecast period. Further, it throws focus on restraints as well discusses future chances at length that are likely to come to the fore over the forecast period. The analysis thus provided helps market stakeholders with business planning and to gauge scope of expansion in the Cognitive Computing market over the forecast period. Moreover, the report has explored changing factors for the market segments. It covers the growth factors of the worldwide market based on end-users. It's a well-crafted Cognitive Computing market research report which has been designed using the primary and secondary sources.

In part 1 of this article we discussed the industry trend of companies wanting to brand themselves as "AI first" and often positioning themselves as deep learning. We highlighted some of the problems building and deploying a deep learning solution presents and suggest that often other machine learning approaches could provide a solution in a simpler and more cost effective way. In this second part we want to outline our own experience building an AI application and reflect on why we chose not to utilise deep learning as the core technology used. At Aiqudo we have built a personal digital assistant for smart phones. Our goal is to understand what users are saying, figure out their intent and execute the correct action for them on their devices..