Highly-trained surgeons armed with a scalpel perform procedures faster than machines, at a lower cost - and do not make more mistakes. Two studies - one by Leeds University and the other by Stanford University in the US - last night independently found robots did not reduce side effects or improve the patient's health when compared to manual operations. But they both found that robotic surgery took longer and was more expensive. For the last 15 years robots have been increasingly used in the NHS, replacing the surgeon's hand with the arm of a machine. The NHS has about 60 surgical robots, often bought for about £1.5million each by hospitals' charitable trusts after local fundraising campaigns.

Over the coming years, the technology company Amazon – like the other Cyber Valley partners from industry – will contribute € 1.25 million to setting up research groups in the Stuttgart and Tübingen region. The Max Planck Society, the federal state of Baden-Württemberg, the universities of Stuttgart and Tübingen as well as companies such as BMW, Bosch, Daimler, IAV, Porsche and ZF Friedrichshafen have joined forces in the collaboration project to drive forward research on intelligent systems and to create an environment for more successful start-ups. The first research groups planned as part of the initiative are currently being set up. "With Amazon joining the Cyber Valley, our idea to create a fruitful environment for business activities will gain momentum by expanding AI research in the Stuttgart and Tübingen area," indicated Martin Stratmann. "Only by bringing together world-class research and entrepreneurial spirit can we create the breeding ground for innovations that may prove to be technological breakthroughs in the future." Other Amazon projects also support the goal of turning Cyber Valley into a creative hotspot for scientific progress and economically successful innovation.

Ha Duong is a former associate at Techstars, product marketing professional, blogger and an aspiring entrepreneur from Germany. One thing I noticed: MMC Ventures is listed twice in "investors_list" column, but I spotted five of their portfolio companies there. I hope my list of European AI startups can help you get an overview of the market and detect new exciting business models you have not heard about. If there is some data you can add or correct please do so and feel free to shoot me a message with thoughts or feedback. Ha Duong, a former associate at Techstars, has compiled a very exhaustive list of European artificial intelligence and machine learning startups.

Data quality and critical thinking are imperative for the advance of artificial intelligence (AI), according to expert views from an event hosted by US-based Teradata. At day three (24 October) of the Teradata Partners Conference 2017 (22-26 October), in Anaheim, California, Banking Technology was there to gauge the zeitgeist in our data-drenched world. As a quick review, on Monday (23 October), the company set out its plans in a super-slick presentation; unveiled Teradata Analytics Platform; revealed Danske Bank has launched its AI driven fraud detection platform; and showed off Agile Analytics Factory. The Era of Human-Machine Collaboration", Stephen Brobst and Yasmeen Ahmad from Teradata; Nadeem Gulzar, Danske Bank; and Andrew Stephen, Said Business School – University of Oxford; discussed a lot more than just coding. Stephen says data quality and critical thinking "should be a given but it's not happening".

Smart cities, which were once confined to the realms of science fiction books, are rapidly becoming a reality all around the globe. Unfortunately, like all revolutionizing innovations, smart cities are developing their own unique challenges alongside of their perks. So what are industry insiders and tomorrow's city planners doing to face these challenges? The security issues facing smart cities are unlike anything ever before seen, and solutions to these problems haven't yet sprung up en masse, meaning many different interest groups have proposed their own respective plans. By combing through some of today's proposed solutions, we can identify some of the leading trends that will come to dominate the future of smart city security.

Professor Bernhard Schölkopf, a director at the Max Planck Institute and leading mind in machine learning, is now also an "Amazon Scholar." Between its online shopping algorithms, online web services and the rise of Alexa, artificial intelligence is becoming an increasingly important part of Amazon's business model. Now, the online mega-retailer is investing in the technology's future with plans for a new Amazon Research Center near the campus of the Max Planck Institute for Intelligent Systems in Tübingen, Germany. Amazon plans to staff the new facility with 100 employees in the field of machine learning, and will pour 1.25 million euros (almost $1.5 million) into new research groups. Research in areas critical to AI, including robotics, machine learning and machine vision, are already underway in the area.

Efficient and high-fidelity prior sampling and inversion for complex geological media is still a largely unsolved challenge. Here, we use a deep neural network of the variational autoencoder type to construct a parametric low-dimensional base model parameterization of complex binary geological media. For inversion purposes, it has the attractive feature that random draws from an uncorrelated standard normal distribution yield model realizations with spatial characteristics that are in agreement with the training set. In comparison with the most commonly used parametric representations in probabilistic inversion, we find that our dimensionality reduction (DR) approach outperforms principle component analysis (PCA), optimization-PCA (OPCA) and discrete cosine transform (DCT) DR techniques for unconditional geostatistical simulation of a channelized prior model. For the considered examples, important compression ratios (200 - 500) are achieved. Given that the construction of our parameterization requires a training set of several tens of thousands of prior model realizations, our DR approach is more suited for probabilistic (or deterministic) inversion than for unconditional (or point-conditioned) geostatistical simulation. Probabilistic inversions of 2D steady-state and 3D transient hydraulic tomography data are used to demonstrate the DR-based inversion. For the 2D case study, the performance is superior compared to current state-of-the-art multiple-point statistics inversion by sequential geostatistical resampling (SGR). Inversion results for the 3D application are also encouraging.

In this paper, a frequency coefficient based on the Sen-Shorrocks-Thon (SST) poverty index notion is proposed. The clustering SST index can be used as the method for determination of the connection between similar neighbor sub-clusters. Consequently, connections can reveal existence of natural homogeneous. Through estimation of the connection strength, we can also verify information about the estimated number of natural clusters that is necessary assumption of efficient market segmentation and campaign management and financial decisions. The index can be used as the complementary tool for the U-matrix visualization. The index is tested on an artificial dataset with known parameters and compared with results obtained by the Unified-distance matrix method.

Performance of data-driven network for tumor classification varies with stain-style of histopathological images. This article proposes the stain-style transfer (SST) model based on conditional generative adversarial networks (GANs) which is to learn not only the certain color distribution but also the corresponding histopathological pattern. Our model considers feature-preserving loss in addition to well-known GAN loss. Consequently our model does not only transfers initial stain-styles to the desired one but also prevent the degradation of tumor classifier on transferred images. The model is examined using the CAMELYON16 dataset.

Data and knowledge representation are fundamental concepts in machine learning. The quality of the representation impacts the performance of the learning model directly. Feature learning transforms or enhances raw data to structures that are effectively exploited by those models. In recent years, several works have been using complex networks for data representation and analysis. However, no feature learning method has been proposed for such category of techniques. Here, we present an unsupervised feature learning mechanism that works on datasets with binary features. First, the dataset is mapped into a feature--sample network. Then, a multi-objective optimization process selects a set of new vertices to produce an enhanced version of the network. The new features depend on a nonlinear function of a combination of preexisting features. Effectively, the process projects the input data into a higher-dimensional space. To solve the optimization problem, we design two metaheuristics based on the lexicographic genetic algorithm and the improved strength Pareto evolutionary algorithm (SPEA2). We show that the enhanced network contains more information and can be exploited to improve the performance of machine learning methods. The advantages and disadvantages of each optimization strategy are discussed.