The asymptotic behavior of the stochastic gradient algorithm with a biased gradient estimator is analyzed. Relying on arguments based on the dynamic system theory (chain-recurrence) and the differential geometry (Yomdin theorem and Lojasiewicz inequality), tight bounds on the asymptotic bias of the iterates generated by such an algorithm are derived. The obtained results hold under mild conditions and cover a broad class of high-dimensional nonlinear algorithms. Using these results, the asymptotic properties of the policy-gradient (reinforcement) learning and adaptive population Monte Carlo sampling are studied. Relying on the same results, the asymptotic behavior of the recursive maximum split-likelihood estimation in hidden Markov models is analyzed, too.

We propose a DTCWT ScatterNet Convolutional Neural Network (DTSCNN) formed by replacing the first few layers of a CNN network with a parametric log based DTCWT ScatterNet. The ScatterNet extracts edge based invariant representations that are used by the later layers of the CNN to learn high-level features. This improves the training of the network as the later layers can learn more complex patterns from the start of learning because the edge representations are already present. The efficient learning of the DTSCNN network is demonstrated on CIFAR-10 and Caltech-101 datasets. The generic nature of the ScatterNet front-end is shown by an equivalent performance to pre-trained CNN front-ends. A comparison with the state-of-the-art on CIFAR-10 and Caltech-101 datasets is also presented.

The heads of more than 100 of the world's top artificial intelligence companies are very alarmed about the development of "killer robots". In an open letter to the UN, these business leaders – including Tesla's Elon Musk and the founders of Google's DeepMind AI firm – warned that autonomous weapon technology could be misused by terrorists and despots or hacked to perform in undesirable ways. But the real threat is much bigger – and not just from human misconduct but from the machines themselves. The research into complex systems shows how behavior can emerge that is much more unpredictable than the sum of individual actions. On one level this means human societies can behave very differently to what you might expect just looking at individual behavior.

Fact Checking Some of the clearest and most objective cases of "Fake News" include incorrect facts that could be checked. This could include lying about the number of immigrants in a country or clearly exaggerated the cost of the national health service for a country. There are more than 114 fact checking initiatives in almost 50 countries. These include Full Fact, a charity in the UK that among other things, fact checks comments made by politicians in the House of Commons and the Crosscheck project from First Draft News that introduces a framework for fact checking that is being tested in the context of the French elections. These organisations are doing an amazing job making sure people are accountable.

Ocado has become the first British supermarket to launch an app on Amazon Alexa meaning customers can add items to their shopping basket using their smart assistant. The service, which launched yesterday, allows customers to order food to be delivered to their homes simply by issuing a voice command. Ocado is the only British supermarket in the UK to offer an app on Alexa, the personal assistant device owned by rival Amazon, as it moves away from traditional supermarket shopping. The service, which launched yesterday, allows customers to order food to be delivered to their homes simply by issuing a voice command. As well as ordering food from the device customers will be able to add a product to an existing order or basket, find out which products are in season and receive inspiration for the best ways to include them in recipes.

Back in January, Google rolled out a "parking difficulty" icon in Maps which, as its name suggests, explains how hard it will be to find a spot for your vehicle. While useful, it was only available in 25 US cities, including New York and San Francisco. Today, it's being expanded to a further 25 locations around the globe: Alicante, Amsterdam, Copenhagen, Barcelona, Cologne, Darmstadt, Dusseldorf, London, Madrid, Malaga, Manchester, Milan, Montreal, Moscow, Munich, Paris, Prague, Rio de Janeiro, Rome, Sao Paulo, Stockholm, Stuttgart, Toronto, Valencia and Vancouver. Just pop in your destination and you'll see a limited, medium or easy symbol near the bottom of the screen, next to the estimated journey time. The ratings are based on "historical parking data," according to Google, and a smidge of "machine learning magic."

Machine learning is rapidly having a big impact on ecommerce as martech and agencies invest in data science. Off the back of this, performance marketing agency iProspect got in touch to tell me about their work with keyword categorisation. I asked a few questions of Owned Media Executive Josh Carty, to find out more. Before we get stuck in, a reminder that this year's Festival of Marketing features one stage (of 12) dedicated to AI in marketing – book your tickets now and see headliners such as Stephen Fry and Jo Malone. Josh Carty: We work with one of the UK's largest online retailers on their organic search and performance content strategy.

With Amazon now a major player in the UK grocery market, supermarkets have been forced to adapt or risk losing some of that all important market share. Some have mimicked their rival by offering one-hour deliveries, while others have attempted to match the online retailer's technological prowess. Already a pioneer of grocery-sorting robots and autonomous deliveries, Ocado has stepped things up a notch, today becoming the first UK supermarket to launch a dedicated Amazon Alexa skill. As you'd expect, the Ocado skill focuses mainly on getting the right products in your basket. But be aware: it's currently not capable of starting a new order, it can only add items to an existing one.

Recruitment systems are evolving so candidates can use artificial intelligence to match themselves to roles in a similar way they might seek out a partner online. This has benefits for employers too, explains Cath Everett. The idea of being able to find the perfect job candidate who is not only ideal in terms of qualifications and experience but also in terms of cultural fit, ethos and personality sounds too good to be true. Technology is disrupting work, but what do staff think? Nonetheless, it is something that a number of artificial intelligence (AI) systems are claiming to offer recruiters.

Spain's agricultural ministry has this month launched, for the second consecutive year, a pilot experiment using artificial intelligence techniques (AI) to anticipate the evolution of the olive fly pest. Having access to predictive information on the behavior of pests is a very powerful tool for technicians and farmers with a view to integrated pest management, since it allows for better decision making, knowing the areas and dates of greater risk and impact, and designing control measures in a more efficient and sustainable manner. To do this, big data collected by the Andalusian Plant Protection and Information Network (RAIF) on this pest and on other crop parameters is analysed and fed into an artificial intelligence model that is able to predict the pest's behaviour up to four weeks in advance, through machine-learning techniques. The European Court of Justice (ECJ) has officially backed drastic measures adopted by the EU to combat a bacteria infection that is decimating Italy's olive crop. The application of "Big Data" management techniques in agriculture, as well as artificial intelligence, are two fields with enormous potential for improving the efficiency and sustainability of the agricultural sector, according to the Spanish agriculture ministry.