During yesterday's Autumn statement, Chancellor Philip Hammond outlined positive measures to push the adoption of autonomous and electric cars, develop new 5G networks, treble the number of computer science teachers and further research into AI and robotics. But tucked away in the 88-page document were small changes that show the UK government plans to get a lot tougher on technology companies that aren't willing to give back as much as they should. The most important notice came during Hammond's budget speech. As he pledged £400 million for a UK-wide EV charging network and a £100 million subsidy for electric car buyers, the finance minister also outlined steps to claw back money from tech giants like Google, Amazon and Apple, which use legal loopholes to avoid paying tax in the UK. "Multinational digital businesses pay billions of pounds in royalties to jurisdictions where they are not taxed – and some of these royalties relate to UK sales," said Hammond in his speech.

Prince Harry and a robot have been announced as two guest editors on Radio 4's Today Programme. Their fellow editors will be Baroness Trumpington, Tamara Rojo and Ben Okri. This is the 14th year control has been handed over to public figures between Christmas and New Year. Kensington Palace said Prince Harry would use the opportunity to "shine a spotlight on issues that are close to his heart". The palace added: "He is working closely with Today's team to produce segments on a range of topics, including youth violence, conservation and mental health."

If you've already paid $60 for a video game, haven't you spent enough? That's the question Electronic Arts, or EA, the maker of games including the Madden NFL series, FIFA and Battlefield, has to answer after angering customers who eagerly anticipated one of its biggest holiday releases, "Star Wars: Battlefront 2." On top of the "Star Wars"-themed action-shooter's $60 list price, the game included micro-transactions, which enabled players to spend real-world money on in-game items such as "loot crates" -- essentially a mystery box filled with perks. Although video games have long allowed players to spend currency on cosmetic purchases such as special costumes, "Battlefront 2" players were upset to learn that a trial version of the game let them spend money to bolster their characters. Those who opted against paying were at a disadvantage and simply had to "grind" -- that is, play for many hours -- to achieve similar powers or unlock marquee characters such as Darth Vader. Players accused EA of engaging in pay-to-win practices.

Machine learning and big data tools similar to those that power popular digital assistants like Alexa and Siri can enable banks and insurance companies to rationalize their operations and cost structures and, longer term, help gain insights about customer needs and identify new sources of incremental revenue. Bank call centers have traditionally been focused on customer satisfaction by responding to routine requests for assistance at minimal cost. They have been run as cost centers, with average call hold time their key metric. Machine learning (ML), Natural Language Processing (NLP) and Robotic Process Automation (RPA) help to develop and automate repetitive tasks and flows. Then Predictive Analytics facilitates building models that are not explicitly programmed.

With each passing month, we see more and more car companies taking a deep dive into artificial intelligence and autonomous systems, as well as studying big data that comes with developing autonomous systems for use in city environments. They do this either by partnering with existing companies or absorbing them, or through loose investments with tech sharing agreements. Audi is starting to train their own employees in-house under the new "data.camp" Despite advances in education and the inclusion of information technology in the most syllabuses around the world, there is still a great number of people in the current workforce that don't quite understand the basics of it. This is especially true in Germany where vocational training means most employees have very narrow ranges of expertise, but with new car development requiring integration with the cloud and such, employees need to understand what they're going to be dealing with.

At home, it helps power personalized shopping apps, suggests personalized entertainment experiences, manages and monitors self-driving cars, supports virtual assistants, and improves navigation. At the office, it helps businesses develop the next best offer, recruit top-notch candidates, detect fraud, automate supply chains, and boost data center efficiency. Yet, corporate finance leaders are asking deeper questions which require more advanced analytics systems. "Why can you ask your mobile phone for directions to find the nearest restaurant but you can't ask your system how revenues are trending in Italy?" Oracle Vice President of Product Strategy for Big Data Analytics, Rich Clayton said during a recent Financial Executives International (FEI) webcast. "Why is that your systems don't understand your processes? Why do you spend so much time explaining simple variances when much can be automated?"

Job Qualifications We are looking for candidates whose teaching and research interests are related to marketing analytics, summarized by one or multiple of the following keywords, amongst statistical and machine learning algorithms, (rule-based/hybrid) ensembles, predictive modeling, R, Python, SAS, Spark, (NO/)SQL, web analytics, web scraping, social media analytics, data mining, recommendation tools, process mining, social network analytics, fraud detection, text mining, visual analytics, and/or big data analysis tools. Applicants should possess a PhD and be able to provide evidence of publications (and/or demonstrate the potential to publish) in reputable academic journals. The candidate will contribute to the IÉSEG Excellence Center for Marketing Analytics and shares his/her expertise within the MSc. in Big Data Analytics for Business. He/she also needs to provide evidence of strong teaching skills and/or professional experience. Applicants should be completely fluent in English as all courses will be taught in this language.

One might think that the BBC would go out of its way to avoid presenters that sound too'robotic'. But now it has appointed an actual robot to present the Today programme, alongside Prince Harry. Experts are using artificial intelligence to create a robotic version of presenter Mishal Husain, which will interview guests on the programme during the Christmas period. The human editors will appear on the Radio 4 show but also get involved behind the scenes, working with staff to decide which topics it should cover. Prince Harry's show will focus on'youth violence, conservation and mental health', whilst Baroness Trumptington will look at'the importance of plain speaking in politics and the debate around the legalisation of brothels'.

To address three important issues involved in latent variable models (LVMs), including capturing infrequent patterns, achieving small-sized but expressive models and alleviating overfitting, several studies have been devoted to "diversifying" LVMs, which aim at encouraging the components in LVMs to be diverse. Most existing studies fall into a frequentist-style regularization framework, where the components are learned via point estimation. In this paper, we investigate how to "diversify" LVMs in the paradigm of Bayesian learning. We propose two approaches that have complementary advantages. One is to define a diversity-promoting mutual angular prior which assigns larger density to components with larger mutual angles and use this prior to affect the posterior via Bayes' rule. We develop two efficient approximate posterior inference algorithms based on variational inference and MCMC sampling. The other approach is to impose diversity-promoting regularization directly over the post-data distribution of components. We also extend our approach to "diversify" Bayesian nonparametric models where the number of components is infinite. A sampling algorithm based on slice sampling and Hamiltonian Monte Carlo is developed. We apply these methods to "diversify" Bayesian mixture of experts model and infinite latent feature model. Experiments on various datasets demonstrate the effectiveness and efficiency of our methods.

Critical periods are phases in the early development of humans and animals during which experience can affect the structure of neuronal networks irreversibly. In this work, we study the effects of visual stimulus deficits on the training of artificial neural networks (ANNs). Introducing well-characterized visual deficits, such as cataract-like blurring, in the early training phase of a standard deep neural network causes irreversible performance loss that closely mimics that reported in humans and animal models. Deficits that do not affect low-level image statistics, such as vertical flipping of the images, have no lasting effect on the ANN's performance and can be rapidly overcome with additional training, as observed in humans. In addition, deeper networks show a more prominent critical period. To better understand this phenomenon, we use techniques from information theory to study the strength of the network connections during training. Our analysis suggests that the first few epochs are critical for the allocation of resources across different layers, determined by the initial input data distribution. Once such information organization is established, the network resources do not re-distribute through additional training. These findings suggest that the initial rapid learning phase of training of ANNs, under-scrutinized compared to its asymptotic behavior, plays a key role in defining the final performance of networks.