We introduce \texttt{pycobra}, a Python library devoted to ensemble learning (regression and classification) and visualisation. Its main assets are the implementation of several ensemble learning algorithms, a flexible and generic interface to compare and blend any existing machine learning algorithm available in Python libraries (as long as a \texttt{predict} method is given), and visualisation tools such as Voronoi tessellations. \texttt{pycobra} is fully \texttt{scikit-learn} compatible and is released under the MIT open-source license. \texttt{pycobra} can be downloaded from the Python Package Index (PyPi) and Machine Learning Open Source Software (MLOSS). The current version (along with Jupyter notebooks, extensive documentation, and continuous integration tests) is available at \href{https://github.com/bhargavvader/pycobra}{https://github.com/bhargavvader/pycobra}.

We introduce and describe the results of a novel shared task on bandit learning for machine translation. The task was organized jointly by Amazon and Heidelberg University for the first time at the Second Conference on Machine Translation (WMT 2017). The goal of the task is to encourage research on learning machine translation from weak user feedback instead of human references or post-edits. On each of a sequence of rounds, a machine translation system is required to propose a translation for an input, and receives a real-valued estimate of the quality of the proposed translation for learning. This paper describes the shared task's learning and evaluation setup, using services hosted on Amazon Web Services (AWS), the data and evaluation metrics, and the results of various machine translation architectures and learning protocols.

Gaussian process (GP) models have become a well-established frameworkfor the adaptive design of costly experiments, and notably of computerexperiments. GP-based sequential designs have been found practicallyefficient for various objectives, such as global optimization(estimating the global maximum or maximizer(s) of a function),reliability analysis (estimating a probability of failure) or theestimation of level sets and excursion sets. In this paper, we dealwith convergence properties of an important class of sequential designapproaches, known as stepwise uncertainty reduction (SUR) strategies.Our approach relies on the key observation that the sequence ofresidual uncertainty measures, in SUR strategies, is generally asupermartingale with respect to the filtration generated by theobservations. We study the existence of SUR strategies and establishgeneric convergence results for a broad class thereof. We alsointroduce a special class of uncertainty measures defined in terms ofregular loss functions, which makes it easier to check that ourconvergence results apply in particular cases. Applications of thelatter include proofs of convergence for the two main SUR strategiesproposed by Bect, Ginsbourger, Li, Picheny and Vazquez (Stat. Comp.,2012). To the best of our knowledge, these are the first convergenceproofs for GP-based sequential design algorithms dedicated to theestimation of excursions sets and their measure. Coming to globaloptimization algorithms, we also show that the knowledge gradientstrategy can be cast in the SUR framework with an uncertaintyfunctional stemming from a regular loss, resulting in furtherconvergence results. We finally establish a new proof of convergencefor the expected improvement algorithm, which is the first proof forthis algorithm that applies to any GP with continuous sample paths.

The statistical dependencies which independent component analysis (ICA) cannot remove often provide rich information beyond the linear independent components. It would thus be very useful to estimate the dependency structure from data. While such models have been proposed, they usually concentrated on higher-order correlations such as energy (square) correlations. Yet, linear correlations are a most fundamental and informative form of dependency in many real data sets. Linear correlations are usually completely removed by ICA and related methods, so they can only be analyzed by developing new methods which explicitly allow for linearly correlated components. In this paper, we propose a probabilistic model of linear non-Gaussian components which are allowed to have both linear and energy correlations. The precision matrix of the linear components is assumed to be randomly generated by a higher-order process and explicitly parametrized by a parameter matrix. The estimation of the parameter matrix is shown to be particularly simple because using score matching, the objective function is a quadratic form. Using simulations with artificial data, we demonstrate that the proposed method improves identifiability of non-Gaussian components by simultaneously learning their correlation structure. Applications on simulated complex cells with natural image input, as well as spectrograms of natural audio data show that the method finds new kinds of dependencies between the components.

The Particle Swarm Optimization Policy (PSO-P) has been recently introduced and proven to produce remarkable results on interacting with academic reinforcement learning benchmarks in an off-policy, batch-based setting. To further investigate the properties and feasibility on real-world applications, this paper investigates PSO-P on the so-called Industrial Benchmark (IB), a novel reinforcement learning (RL) benchmark that aims at being realistic by including a variety of aspects found in industrial applications, like continuous state and action spaces, a high dimensional, partially observable state space, delayed effects, and complex stochasticity. The experimental results of PSO-P on IB are compared to results of closed-form control policies derived from the model-based Recurrent Control Neural Network (RCNN) and the model-free Neural Fitted Q-Iteration (NFQ). Experiments show that PSO-P is not only of interest for academic benchmarks, but also for real-world industrial applications, since it also yielded the best performing policy in our IB setting. Compared to other well established RL techniques, PSO-P produced outstanding results in performance and robustness, requiring only a relatively low amount of effort in finding adequate parameters or making complex design decisions.

A new week brings a fresh Google initiative targeting AI startups. We started the month with the announcement of Gradient Ventures, Google's on-balance sheet AI investment vehicle. Two days later we watched the finalists of Google Cloud's machine learning competition pitch to a panel of top AI investors. And today, Google's Launchpad is announcing a new hands-on Studio program to feed hungry AI startups the resources they need to get off the ground and scale. The thesis is simple -- not all startups are created the same.

Passengers at all U.S. airports will soon face new security measures for their tablets, e-readers and video game consoles. Transportation Security Administration (TSA) officers will order travelers to take all devices larger than a cellphone out of their bag and put them in a bin by themselves. Prior rules required only laptops to be removed for separate screening. Officials say it gives X-ray screeners a clearer picture of the devices. Passengers at all U.S. airports will soon face new security measures for their tablets, e-readers and video game consoles (stock image) TSA said the new rules have been in place in a pilot project at ten American airports and will expand to all US airports in the months ahead.

WASHINGTON – The U.S. Transportation Security Administration (TSA) said Wednesday it will impose new stricter security rules requiring airline travelers to remove all electronic items larger than mobile phones, including tablets, e-readers and video game consoles, from carry-on baggage for screening. Prior rules required only laptops to be removed for separate screening. The new rules significantly expand the number of electronic devices that will need to be removed for screening and help government employees get a clearer view during X-ray screening. TSA said the new rules have been in place in a pilot project at 10 U.S. airports, including Detroit, Los Angeles, Boston and Phoenix, and will expand to all U.S. airports in the months ahead. The new enhanced security rules at U.S. airports only apply at standard security lanes -- not at lanes for travelers who are in "pre-check" programs.

As it is typical for any realpolitik, citizens are becoming less relevant in digital realpolitik. They are personally targeted in advertising and surveillance efforts by corporations and governments. Individuals per se are getting lost in big numbers. The individual is just one amongst billions of Facebook users, and just one amongst billions of contributors to Google searches. Governments are increasingly speaking about digital sovereignty and less about the empowerment of individuals. Citizens are becoming more and more the object of digital growth and less and less the engine behind it, as it has been since the early days of the Internet. On a promising note, realpolitik provides a more realistic picture of interests and risks as well as winners and losers resulting from Internet developments. It is in this way that realpolitik can contribute to creating the basis for more solid and sustainable Internet development. Governments are likely to continue striking deals with Internet companies in order to recuperate some taxes. The bilateral deals could be the building blocks for a more structured approach to revenues from the digital economy.

A new week brings a fresh Google initiative targeting AI startups. We started the month with the announcement of Gradient Ventures, Google's on-balance sheet AI investment vehicle. Two days later we watched the finalists of Google Cloud's machine learning competition pitch to a panel of top AI investors. And today, Google's Launchpad is announcing a new hands-on Studio program to feed hungry AI startups the resources they need to get off the ground and scale. The thesis is simple -- not all startups are created the same.