Through 5G related research, the door to the so called millimeter wave (mmWave) frequencies reopened, unlocking a huge chunk of untapped bandwidth [1]. With mmWaves, the propagation changes dramatically: the resulting radiation has severe path loss properties and reflects on most visible obstacles [2]. To counteract the aforementioned characteristics, beamforming (BF) is usually employed in systems containing multiple-input and multiple-output (MIMO) antennas, enabling steerable and focused radiation patterns. With that recent focus on mmWaves, new positioning systems based on these frequencies were proposed [3]. The achievable accuracy in controlled conditions is remarkable, with sub-meter accuracy in indoor [4] and ultra-dense line-ofsight (LOS) outdoor scenarios [5]. Nevertheless, in order to be useful in outdoor scenarios, a mmWave positioning system must also be able to deal with devices in non-line-of-sight (NLOS) locations. The works developed in [6]-[9] attempt to address this concern, being capable of locating devices in both LOS and NLOS situations. The method in [6] applies compressed sensing on information gathered from static listeners, while in [7] multiple access points are used to create a location fingerprint database of received powers and angles-of-arrival (AoA). In [8], the authors use multiple BF transmissions and an iterative algorithm to estimate the position and orientation of the device.

Skip-gram with negative sampling, a popular variant of Word2vec originally designed and tuned to create word embeddings for Natural Language Processing, has been used to create item embeddings with successful applications in recommendation. While these fields do not share the same type of data, neither evaluate on the same tasks, recommendation applications tend to use the same already tuned hyperparameters values, even if optimal hyperparameters values are often known to be data and task dependent. We thus investigate the marginal importance of each hyperparameter in a recommendation setting, with an extensive joint hyperparameter optimization on various datasets. Results reveal that optimizing neglected hyperparameters, namely negative sampling distribution, number of epochs, subsampling parameter and window-size, significantly improves performance on a recommendation task, and can increase it up to a factor of $10$.

Existing supervised approaches didn't make use of the low-level features which are actually effective to this task. And another deficiency is that they didn't consider the relation between pixels, which means effective features are not extracted. In this paper, we proposed a novel convolutional neural network which make sufficient use of low-level features together with high-level features and involves atrous convolution to get multi-scale features which should be considered as effective features. Our model is tested on three standard benchmarks - DRIVE, STARE, and CHASE databases. The results presents that our model significantly outperforms existing approaches in terms of accuracy, sensitivity, specificity, the area under the ROC curve and the highest prediction speed. Our work provides evidence of the power of wide and deep neural networks in retinal blood vessels segmentation task which could be applied on other medical images tasks.

Computer vision systems for automatic image categorization have become accurate and reliable enough that they can run continuously for days or even years as components of real-world commercial applications. A major open problem in this context, however, is quality control. Good classification performance can only be expected if systems run under the specific conditions, in particular data distributions, that they were trained for. Surprisingly, none of the currently used deep network architectures has a built-in functionality that could detect if a network operates on data from a distribution that it was not trained for and potentially trigger a warning to the human users. In this work, we describe KS(conf), a procedure for detecting such outside of the specifications operation. Building on statistical insights, its main step is the applications of a classical Kolmogorov-Smirnov test to the distribution of predicted confidence values. We show by extensive experiments using ImageNet, AwA2 and DAVIS data on a variety of ConvNets architectures that KS(conf) reliably detects out-of-specs situations. It furthermore has a number of properties that make it an excellent candidate for practical deployment: it is easy to implement, adds almost no overhead to the system, works with all networks, including pretrained ones, and requires no a priori knowledge about how the data distribution could change. 1 Introduction With the emergence of deep convolutional networks (ConvNets), computer vision systems have become accurate and reliable enough to perform tasks of practical relevance autonomously and reliably over long periods of time. This work was in parts funded by the European Research Council under the European Unions Seventh Framework Programme (FP7/2007-2013)/ERC grant agreement no 308036. C. H. Lampert, IST Austria, Email: chl@ist.ac.at 2 Rémy Sun, Christoph H. Lampert "ski" "shovel" "web site" "tennis ball" Figure 1 Illustration of within specification and outside of specifications behavior of a Conv-Net (here: VGG19, trained on ILSVRC2012). Left image: prediction on images that the network was trained to recognize. We observe: a standard multi-class network always predicts one of its predefined class labels, even if the current input is distorted, or even completely different, from what it was trained for. A major concern in our society about automatic decision systems is their reliability: if decisions are made by a trained classifier instead of a person, how can we be sure that the system works reliably now, and that it will continue to do so in the future?

Market making is a fundamental trading problem in which an agent provides liquidity by continually offering to buy and sell a security. The problem is challenging due to inventory risk, the risk of accumulating an unfavourable position and ultimately losing money. In this paper, we develop a high-fidelity simulation of limit order book markets, and use it to design a market making agent using temporal-difference reinforcement learning. We use a linear combination of tile codings as a value function approximator, and design a custom reward function that controls inventory risk. We demonstrate the effectiveness of our approach by showing that our agent outperforms both simple benchmark strategies and a recent online learning approach from the literature.

This article is inspired by the research done during studies in the university. Goal of this article is to act as a note and reminder that finding anomalies is not a trivial task (currently). Anomaly detection refers to the task of finding observations that do not conform to the normal, expected behaviour. These observations can be named as anomalies, outliers, novelty, exceptions, surprises in different application domains. The most popular terms that occur most often in literature are anomalies and outliers.

Speaking at the Sorbonne in September 2017, French president Emmanuel Macron made clear his ambitions for Europe to become a global leader in the field of artificial intelligence (AI). The European Commission is presently working on a strategy on this and is set to deliver a communication on it in the coming months. But there is much uncertainty over how ambitions such as Macron's can be attained. Much will depend on how the European Court of Justice and member states interpret a number of key provisions of the General Data Protection Regulation (GDPR). This is the mammoth culmination of the European Union's five-year effort to make European data-protection law fit for the 21st century.

A NEW commodity spawns a lucrative, fast-growing industry, prompting antitrust regulators to step in to restrain those who control its flow. A century ago, the resource in question was oil. Now similar concerns are being raised by the giants that deal in data, the oil of the digital era. These titans--Alphabet (Google's parent company), Amazon, Apple, Facebook and Microsoft--look unstoppable. They are the five most valuable listed firms in the world.

Leading South African law firm, Webber Wentzel, has announced it has chosen legal AI company Luminance to provide doc review services for M&A transactions. This is the latest in a series of client wins for the UK-based AI company, which last week also announced it had partnered with Sweden's Delphi and also Luxembourg's Arendt & Medernach. In the latter case the Benelux firm will be making use of Luminance's new real estate document review capacity. Until recently the company had been focused on M&A due diligence work. In a statement Webber Wentzel said it'particularly values the platform's built-in collaboration tools which will allow its lawyers to quickly group and assign documents, track live progress, and significantly reduce the amount of time spent organising workflow'.

Trust us, we have a team of hardcore growth hackers and Artificial Intelligence professionals who know how to pick the real talent. You start by filling out our questionnaire. When you fit our selection criteria Emma will get in touch with you to plan in a Skype call. If we are still convinced you could be one of the trainees, we will send you a Challenge and let you run some tests. You will have a second Skype call and to discuss the results.