Businesses in general recognize the critical need to believe in, and invest in, disruptive, innovative technologies. Yet many are plagued by inertia and indecision. Many banks certainly suffer from this pair of troubles. However, it also seems that the banking industry, as a whole, is breaking out of tech ennui in a very particular way--by adopting artificial intelligence applications. On the downer side, a survey sponsored by Dell EMC and conducted by Enterprise Strategy Group collected responses from 1,000 executives at a variety of large global companies.

Artificial intelligence (AI) will "significantly transform" the insurance industry in the next three years, with insurers investing in AI to empower agents, brokers and employees to enhance the customer experience with automated personalized services, faster claims handling and individual risk-based underwriting processes, according to a new report from Accenture. The Technology Vision for Insurance 2017 report, called Technology for People, released on Wednesday by the global professional services company, found that while the technology will be empowering, insurers face challenges integrating AI into their existing technology. Insurers cite issues such as data quality, privacy and infrastructure compatibility. The report is based on the insights of a technology advisory board, interviews with industry technologists and a survey of more than 550 insurance executives across 31 countries in North America, Europe, Asia-Pacific, Africa and South America, Accenture noted in a press release. The goal of the survey was to identify the key issues and priorities for technology adoption and investment.

This article is a gentle introduction to attentional and memory-based interfaces in deep neural architectures, using TensorFlow. Incorporating attention mechanisms is very simple and can offer transparency and interpretability to our complex models. We conclude with extensions and caveats of the interfaces. As you read the article, please access all of the code on GitHub and view the IPython notebook here; all code is compatible with TensorFlow version 1.0. The intended audience for this notebook are developers and researchers who have some basic understanding of TensorFlow and fundamental deep learning concepts.

Consumers aren't convinced that AI can provide the same, or better level of customer service that a human can provide today, according to a Pegasystems survey. In fact, more adults surveyed across the globe disagree (38%) than agree (27%) that AI can match human customer service. However, there's a bit more optimism about the future, as more agree (38%) than disagree (26%) that AI has the potential to improve service in the future. Even so, more than one-third are unsure, suggesting that there's some apathy towards the subject. The results of the survey show that more people prefer to talk to a live representative on the phone (45%) than any other method of contact for customer service.

In recent years, we have become increasingly good at training deep neural networks to learn a very accurate mapping from inputs to outputs, whether they are images, sentences, label predictions, etc. from large amounts of labeled data. What our models still frightfully lack is the ability to generalize to conditions that are different from the ones encountered during training. Every time you apply your model not to a carefully constructed dataset but to the real world. The real world is messy and contains an infinite number of novel scenarios, many of which your model has not encountered during training and for which it is in turn ill-prepared to make predictions. The ability to transfer knowledge to new conditions is generally known as transfer learning and is what we will discuss in the rest of this post. Over the course of this blog post, I will first contrast transfer learning with machine learning's most pervasive and successful paradigm, supervised learning. I will then outline reasons why transfer learning warrants our attention. Subsequently, I will give a more technical definition and detail different transfer learning scenarios.

In his session at @ThingsExpo, Eric Lachapelle, CEO of the Professional Evaluation and Certification Board (PECB), will provide an overview of various initiatives to certifiy the security of connected devices and future trends in ensuring public trust of IoT. Speaker Bio Eric Lachapelle is the Chief Executive Officer of the Professional Evaluation and Certification Board (PECB), an international certification body. His role is to help companies and individuals to achieve professional, accredited and worldwide recognized certification against various international standards. He also has extensive experience as a trainer and an educator in the fields of Information Security, Risk Management and IT. Throughout his career, he has worked in North America, Latin America and Asia with individuals and various companies of all sizes.

The first thought is about open sourcing technologies. I have already written on this trend, which is quite unusual at a first look if you think about it, but my thinking around open source has been highly stimulated by the talk given by Wes McKinney -- for who doesn't know who he is, well, he is definitely not a random guy but is THE open source guy (creator of pandas and author of Python for Data Analysis). The open source model is quite hard to be reconciled with the traditional SaaS model, especially in the financial sector. However, we are observing many firms providing cutting-edge technologies and algorithms for free. While in some cases there is a specific business motivation behind it (e.g., Google releasing Tensorflow to avoid conflict of interests with their cloud offering), the decision of open sourcing (part of) the technology actually represents an emerging trend.

Reinforcement learning is considered to be a strong AI paradigm which can be used to teach machines through interaction with the environment and learning from their mistakes. Despite its perceived utility, it has not yet been successfully applied in automotive applications. Motivated by the successful demonstrations of learning of Atari games and Go by Google DeepMind, we propose a framework for autonomous driving using deep reinforcement learning. This is of particular relevance as it is difficult to pose autonomous driving as a supervised learning problem due to strong interactions with the environment including other vehicles, pedestrians and roadworks. As it is a relatively new area of research for autonomous driving, we provide a short overview of deep reinforcement learning and then describe our proposed framework. It incorporates Recurrent Neural Networks for information integration, enabling the car to handle partially observable scenarios. It also integrates the recent work on attention models to focus on relevant information, thereby reducing the computational complexity for deployment on embedded hardware. The framework was tested in an open source 3D car racing simulator called TORCS. Our simulation results demonstrate learning of autonomous maneuvering in a scenario of complex road curvatures and simple interaction of other vehicles.

It is approximately 50 years since the first computational experiments were conducted in what has become known today as the field of Genetic Programming (GP), twenty years since John Koza named and popularised the method, and ten years since the first issue appeared of the Genetic Programming & Evolvable Machines journal. In particular, during the past two decades there has been a significant range and volume of development in the theory and application of GP, and in recent years the field has become increasingly applied. There remain a number of significant open issues despite the successful application of GP to a number of challenging real-world problem domains and progress in the development of a theory explaining the behavior and dynamics of GP. These issues must be addressed for GP to realise its full potential and to become a trusted mainstream member of the computational problem solving toolkit. In this paper we outline some of the challenges and open issues that face researchers and practitioners of GP.

In many sequential decision-making problems one is interested in minimizing an expected cumulative cost while taking into account \emph{risk}, i.e., increased awareness of events of small probability and high consequences. Accordingly, the objective of this paper is to present efficient reinforcement learning algorithms for risk-constrained Markov decision processes (MDPs), where risk is represented via a chance constraint or a constraint on the conditional value-at-risk (CVaR) of the cumulative cost. We collectively refer to such problems as percentile risk-constrained MDPs. Specifically, we first derive a formula for computing the gradient of the Lagrangian function for percentile risk-constrained MDPs. Then, we devise policy gradient and actor-critic algorithms that (1) estimate such gradient, (2) update the policy in the descent direction, and (3) update the Lagrange multiplier in the ascent direction. For these algorithms we prove convergence to locally optimal policies. Finally, we demonstrate the effectiveness of our algorithms in an optimal stopping problem and an online marketing application.