Over 6,000 people are attending a conference focusing on artificial intelligence (AI) that opened in Amsterdam this morning. World Summit AI brings together corporates, startups, investors, scientists, academics, NGOs along with government bodies like the UN, EU and the World Economic Forum. Participants will be learning about some of the latest innovations in AI – the creation of human-like technology - that will transform business and the ethical issues that come with it. The event coincides with Artificial Intelligence in Europe, a report by Microsoft that reveals over half of the companies surveyed expect AI to have an impact on "business areas that are entirely unknown today". Yet only 4% of companies actively use AI suggesting European businesses, at least, have an enormous mountain to climb.

Rogue drones have nearly caused air accidents, have been used as offensive weapons, to deliver drugs to prisoners, and to spy on people. So how can we fight back? This summer a packed Airbus A321 came within 100ft (30m) of disaster after encountering a drone at 15,500ft. And the number of near-misses of this sort has trebled over the last three years, with 92 incidents reported last year in the UK alone. Dozens were classified as involving a serious chance of a collision.

In 2014, we wrote about some failsafe software from ETH Zurich that allowed a quadrotor to remain fully controllable even with one busted motor. The unbalanced torque generated by three motors means that a quadrotor can't help but spin, but with a bit of cleverness, software can compensate for the spin and keep the quadrotor stable and even allow it to obey control inputs, allowing it to land more or less safely. This is a valuable capability, but there are a few things that it doesn't address. For example, what if your quadrotor loses a rotor over an unsafe area? What if something happens to it when it's already traveling at a high speed?

Complex industrial systems are continuously monitored by a large number of heterogenous sensors. The diversity of their operating conditions and the possible fault types make it impossible to collect enough data for learning all the possible fault patterns. The paper proposes an integrated automatic unsupervised feature learning approach for fault detection that uses healthy conditions data only for its training. The approach is based on stacked Extreme Learning Machines (namely Hierarchical, or HELM) and comprises stacked autoencoders performing unsupervised feature learning, and a one-class classifier monitoring the variations in the features to assess the health of the system. This study provides a comprehensive evaluation of HELM fault detection capability compared to other machine learning approaches, including Deep Belief Networks. The performance is first evaluated on a synthetic dataset with typical characteristics of condition monitoring data. Subsequently, the approach is evaluated on a real case study of a power plant fault. HELM demonstrates a better performance specifically in cases where several non-informative signals are included.

We develop model-based methods for solving stochastic convex optimization problems, introducing the approximate-proximal point, or \aProx, family, which includes stochastic subgradient, proximal point, and bundle methods. When the modeling approaches we propose are appropriately accurate, the methods enjoy stronger convergence and robustness guarantees than classical approaches, even though the model-based methods typically add little to no computational overhead over stochastic subgradient methods. For example, we show that improved models converge with probability 1 and enjoy optimal asymptotic normality results under weak assumptions; these methods are also adaptive to a natural class of what we term easy optimization problems, achieving linear convergence under appropriate strong growth conditions on the objective. Our substantial experimental investigation shows the advantages of more accurate modeling over standard subgradient methods across many smooth and non-smooth optimization problems.

The Universal Turing Machine (TM) is a model for VonNeumann computers --- general-purpose computers. A human brain can inside-skull-automatically learn a universal TM so that he acts as a general-purpose computer and writes a computer program for any practical purposes. It is unknown whether a machine can accomplish the same. This theoretical work shows how the Developmental Network (DN) can accomplish this. Unlike a traditional TM, the TM learned by DN is a super TM --- Grounded, Emergent, Natural, Incremental, Skulled, Attentive, Motivated, and Abstractive (GENISAMA). A DN is free of any central controller (e.g., Master Map, convolution, or error back-propagation). Its learning from a teacher TM is one transition observation at a time, immediate, and error-free until all its neurons have been initialized by early observed teacher transitions. From that point on, the DN is no longer error-free but is always optimal at every time instance in the sense of maximal likelihood, conditioned on its limited computational resources and the learning experience. This letter also extends the Church-Turing thesis to automatic programming for general purposes and sketchily proved it.

Meta-data from photo-sharing websites such as Flickr can be used to obtain rich bag-of-words descriptions of geographic locations, which have proven valuable, among others, for modelling and predicting ecological features. One important insight from previous work is that the descriptions obtained from Flickr tend to be complementary to the structured information that is available from traditional scientific resources. To better integrate these two diverse sources of information, in this paper we consider a method for learning vector space embeddings of geographic locations. We show experimentally that this method improves on existing approaches, especially in cases where structured information is available.

Understanding how a learned black box works is of crucial interest for the future of Machine Learning. In this paper, we pioneer the question of the global interpretability of learned black box models that assign numerical values to symbolic sequential data. To tackle that task, we propose a spectral algorithm for the extraction of weighted automata (WA) from such black boxes. This algorithm does not require the access to a dataset or to the inner representation of the black box: the inferred model can be obtained solely by querying the black box, feeding it with inputs and analyzing its outputs. Experiments using Recurrent Neural Networks (RNN) trained on a wide collection of 48 synthetic datasets and 2 real datasets show that the obtained approximation is of great quality.

Clustering non-Euclidean data is difficult, and one of the most used algorithms besides hierarchical clustering is the popular algorithm PAM, partitioning around medoids, also known as k-medoids. In Euclidean geometry the mean--as used in k-means--is a good estimator for the cluster center, but this does not hold for arbitrary dissimilarities. PAM uses the medoid instead, the object with the smallest dissimilarity to all others in the cluster. This notion of centrality can be used with any (dis-)similarity, and thus is of high relevance to many domains such as biology that require the use of Jaccard, Gower, or even more complex distances. A key issue with PAM is, however, its high run time cost. In this paper, we propose modifications to the PAM algorithm where at the cost of storing O(k) additional values, we can achieve an O(k)-fold speedup in the second ("SWAP") phase of the algorithm, but will still find the same results as the original PAM algorithm. If we slightly relax the choice of swaps performed (while retaining comparable quality), we can further accelerate the algorithm by performing up to k swaps in each iteration. We also show how the CLARA and CLARANS algorithms benefit from this modification. In experiments on real data with k=100, we observed a 200 fold speedup compared to the original PAM SWAP algorithm, making PAM applicable to larger data sets, and in particular to higher k.

Assessing the predictive accuracy of black box classifiers is challenging in the absence of labeled test datasets. In these scenarios we may need to rely on a human oracle to evaluate individual predictions; presenting the challenge to create query algorithms to guide the search for points that provide the most information about the classifier's predictive characteristics. Previous works have focused on developing utility models and query algorithms for discovering unknown unknowns --- misclassifications with a predictive confidence above some arbitrary threshold. However, if misclassifications occur at the rate reflected by the confidence values, then these search methods reveal nothing more than a proper assessment of predictive certainty. We are unable to properly mitigate the risks associated with model deficiency when the model's confidence in prediction exceeds the actual model accuracy. We propose a facility locations utility model and corresponding greedy query algorithm that instead searches for overconfident unknown unknowns. Through robust empirical experiments we demonstrate that the greedy query algorithm with the facility locations utility model consistently results in oracle queries with superior performance in discovering overconfident unknown unknowns than previous methods.