But if you're looking for our most hated automated thing in the universe, that's easy. It's the Interactive Voice Response (IVR) system, better known as a phone tree. While bots may one day have the potential to replace customer support and e-commerce, human touch will remain essential to the experience for the foreseeable future. Leverage bots to handle the basics, but also create easy opportunities for customers to connect with a live human (or have one call back) if the interaction gets stalled.

Accenture have announced a partnership with IPsoft, the creators of the cognitive agent Amelia. The two firms will collaborate to create the first Accenture Amelia practice, designed to accelerate client adoption of artificial intelligence to improve business outcomes and create new growth opportunities for their businesses. Accenture will utilize IPsoft's Amelia platform to develop go-to-market strategies, solutions and consulting service offerings around deployments of virtual agent technology for clients across several industries with initial focus on banking, insurance and travel. "Artificial intelligence is maturing rapidly and offers great potential to reshape the way that organizations conduct business and interact with their customers and employees. At the same time, executives are overwhelmed by the plethora of technologies and many products that are advertising AI or Cognitive capabilities," said Paul Daugherty, Accenture's chief technology officer.

Man Group is changing the focus of an Oxford University department it funds to machine learning. Investors have been urged to reject its pay plans later this month. City Talk: Barratt Developments sees strong market, Compass profit rises, Experian plans share buyback. Bank league tables can be a surprisingly thorny issue. Cut the right way, anyone can be number one.

Identifying the unknown underlying trend of a given noisy signal is extremely useful for a wide range of applications. The number of potential trends might be exponential, which can be computationally exhaustive even for short signals. Another challenge, is the presence of abrupt changes and outliers at unknown times which impart resourceful information regarding the signal's characteristics. In this paper, we present the $\ell_1$ Adaptive Trend Filter, which can consistently identify the components in the underlying trend and multiple level-shifts, even in the presence of outliers. Additionally, an enhanced coordinate descent algorithm which exploit the filter design is presented. Some implementation details are discussed and a version in the Julia language is presented along with two distinct applications to illustrate the filter's potential.

Research indicates that IoT and Machine Learning are more valuable to utilities when used in combination but there are hurdles to overcome first. Machine learning and IoT will enable utilities to better realize the next generation of the power grid: a distributed system with power flows among millions of things like distributed energy resources (DERs), microgrids and in-home devices. All of which will help utilities deliver clean reliable energy and greater customer choice. Utility respondents to new research from SAS and Zpryme, The Autonomous Grid, indicated that IoT and machine learning were more than market hype. These technologies are already delivering actionable results, say respondents.

It's a strong contender for the geekiest video ever made: a close-up of a smartphone with line upon line of numbers and symbols scrolling down the screen. But when visitors stop by Nicola Marzari's office, which overlooks Lake Geneva, he can hardly wait to show it off. "It's from 2010," he says, "and this is my cellphone calculating the electronic structure of silicon in real time!" Even back then, explains Marzari, a physicist at the Swiss Federal Institute of Technology in Lausanne (EPFL), Switzerland, his now-ancient handset took just 40 seconds to carry out quantum-mechanical calculations that once took many hours on a supercomputer -- a feat that not only shows how far such computational methods have come in the past decade or so, but also demonstrates their potential for transforming the way materials science is done in the future. Instead of continuing to develop new materials the old-fashioned way -- stumbling across them by luck, then painstakingly measuring their properties in the laboratory -- Marzari and like-minded researchers are using computer modelling and machine-learning techniques to generate libraries of candidate materials by the tens of thousands.

The backpropagation algorithm for calculating gradients has been widely used in computation of weights for deep neural networks (DNNs). This method requires derivatives of objective functions and has some difficulties finding appropriate parameters such as learning rate. In this paper, we propose a novel approach for computing weight matrices of fully-connected DNNs by using two types of semi-nonnegative matrix factorizations (semi-NMFs). In this method, optimization processes are performed by calculating weight matrices alternately, and backpropagation (BP) is not used. We also present a method to calculate stacked autoencoder using a NMF. The output results of the autoencoder are used as pre-training data for DNNs. The experimental results show that our method using three types of NMFs attains similar error rates to the conventional DNNs with BP.

Young patients at Alder Hey will soon be able to use their smartphones to ask questions about everything from hospital menus to details of their medical care. The West Derby hospital has teamed up with computer giant IBM to use artificial intelligence to help put children and their families at ease. The technology – known as Watson – will allow kids to ask questions ahead of their admission to hospital, relaxing them as much as possible beforehand. Over the next few months, hundreds of Alder Hey patients and their parents will be asked a range of questions about everything from parking to what they would like to eat, their favourite games and films and what they want their bedroom to look like.

A large number of the presentations at ARC Advisory Group's recent Industry Forum in Orlando, Florida, focused on how Industrial Internet of Things (IIoT) technologies, such as smart sensors, predictive analytics, and machine learning, can be applied to improve the availability, reliability, and performance of industrial assets and enable new business models. In one presentation, Rob Miller, General Manager, Global Solutions for Flowserve, discussed how the company plans a step change in maintenance and reliability practices for its customers by integrating advanced machine learning capabilities. Flowserve has been evaluating the potential of machine learning to improve its equipment monitoring capabilities for the past twenty years, but – until recently – these had proven too costly and difficult to commercialize. As Mr. Miller explained, Flowserve's initial approach involved increasing data acquisition capabilities utilizing wireless technologies to bring pump health data up to the plant-, or cloud-level and eventually migrate to actively monitoring its customers' equipment. However, with this approach, the company faced many challenges in predicting equipment failures with adequate advanced notice to allow its customers to react effectively to alerts.

Exploration-exploitation of functions, that is learning and optimizing a mapping between inputs and expected outputs, is ubiquitous to many real world situations. These situations sometimes require us to avoid certain outcomes at all cost, for example because they are poisonous, harmful, or otherwise dangerous. We test participants' behavior in scenarios in which they have to find the optimum of a function while at the same time avoid outputs below a certain threshold. In two experiments, we find that Safe-Optimization, a Gaussian Process-based exploration-exploitation algorithm, describes participants' behavior well and that participants seem to care firstly whether a point is safe and then try to pick the optimal point from all such safe points. This means that their trade-off between exploration and exploitation can be seen as an intelligent, approximate, and homeostasis-driven strategy.