It is not every day you see a runaway robot causing traffic chaos in a city centre - but in one Russian suburb is has happened twice in the last week. The robot, named Promobot, was being put through its paces at a research lab in the city of Perm in central Russia's Perm Krai region. In its first escape, the robot, designed to avoid obstacles and to turn around when it reached a boundary, had been left walking around an outside yard. This is the hilarious moment a runaway robot causes traffic chaos in a city centre. The robot - called Promobot - was being put through its paces at a research lab in the city of Perm in central Russia's Perm Krai region Promobot - short for Promotional Robot - is a unique robot created by Russian scientists and is designed to work in customer relations.

Fully self-driving cars are still a few years away, but Mercedes-Benz makes a leap forward with features in the new 2017 E-Class, its midsize luxury sedan and one of its best-selling models. All the basic stuff you'd expect in a German luxury car is there, such as adaptive cruise control, plenty of airbags, and the system that helps protect passengers if the car determines an accident is imminent, tightening seat belts and aggressively braking the car. This feature includes a lane change feature. Once you've set the cruise control, you can use the turn signals to make the car change lanes without your intervention. It will make the change as long as the car doesn't detect a vehicle in the adjacent lane or one approaching at a high rate of speed, which it can determine using its long-range rear-facing radar.

Pedro Domingos speaks on the future of the Information Age. Machine learning not only affects computers, but it will also change our lives. Pedro asks "what will the ultimate learning algorithm look like?" and discusses how future technology will change how we model many parts of our lives. Pedro Domingos is a professor of computer science at the University of Washington and the author of "The Master Algorithm". He is a winner of the SIGKDD Innovation Award, the highest honor in data science.

Charles Radclyffe, a serial entrepreneur who has focused his career on solving tough technology challenges for some of the world's largest organisations, spoke about data philosophy and mentioned how technology is slowly changing the world and could cause a wealth distribution imbalance. The Big Data Summit (Malta) is the first event of its kind to be held in Malta aimed at bringing together an international group of business leaders, policy makers and technology leaders to discuss the future of the global economy and how big data and advanced analytics is already transforming the business world as we know it. This major event brings together thought leaders from some of the key players in Big Data today including Tableau, Qlik, Microsoft, Zendesk and Salesforce as well as accomplished independent international speakers from a variety of industries, including professional services, IT, Telco, iGaming, Academia as well as areas where some of the major breakthroughs are being made like Machine Learning and Artificial Intelligence. Mr Radclyffe spoke about data philosophy and data ethics. He said that the consequence of what we are doing through technology today will have the furthest reaching impact to date.

This week, the U.S. Federal Aviation Administration announced new safety regulations for unmanned aircraft weighing less than 55 pounds (25 kilograms) that are conducting non-hobbyist operations. In other words, the pilots and drones shooting your wedding video, trailing a snowboarder to catch the best trick as seen from above, or taking aerial footage of the horse ranch for sale in the next county now have dictates to follow. The general sUAS (small unmanned aircraft systems) rules that the FAA announced last year did a reasonable job of regulating small drones flown by hobbyists for fun. However, the rules did not make life any easier for anyone who wanted to fly a drone while making money; commercial operators were still required to register separately through a cumbersome and antiquated process involving paper. The FAA promised that sometime in the spring of this year, they'd announce a streamlined registration process for commercial sUAS.

The analysis and diagnosis of Alzheimer's disease (AD) can be based on genetic variations, e.g., single nucleotide polymorphisms (SNPs) and phenotypic traits, e.g., Magnetic Resonance Imaging (MRI) features. We consider two important and related tasks: i) to select genetic and phenotypical markers for AD diagnosis and ii) to identify associations between genetic and phenotypical data. While previous studies treat these two tasks separately, they are tightly coupled because underlying associations between genetic variations and phenotypical features contain the biological basis for a disease. Here we present a new sparse Bayesian approach for joint association study and disease diagnosis. In this approach, common latent features are extracted from different data sources based on sparse projection matrices and used to predict multiple disease severity levels; in return, the disease status can guide the discovery of relationships between data sources. The sparse projection matrices not only reveal interactions between data sources but also select groups of biomarkers related to the disease. Moreover, to take advantage of the linkage disequilibrium (LD) measuring the non-random association of alleles, we incorporate a graph Laplacian type of prior in the model. To learn the model from data, we develop an efficient variational inference algorithm. Analysis on an imaging genetics dataset for the study of Alzheimer's Disease (AD) indicates that our model identifies biologically meaningful associations between genetic variations and MRI features, and achieves significantly higher accuracy for predicting ordinal AD stages than the competing methods.

Process mining techniques aim to extract insights in processes from event logs. One of the challenges in process mining is identifying interesting and meaningful event labels that contribute to a better understanding of the process. Our application area is mining data from smart homes for elderly, where the ultimate goal is to signal deviations from usual behavior and provide timely recommendations in order to extend the period of independent living. Extracting individual process models showing user behavior is an important instrument in achieving this goal. However, the interpretation of sensor data at an appropriate abstraction level is not straightforward. For example, a motion sensor in a bedroom can be triggered by tossing and turning in bed or by getting up. We try to derive the actual activity depending on the context (time, previous events, etc.). In this paper we introduce the notion of label refinements, which links more abstract event descriptions with their more refined counterparts. We present a statistical evaluation method to determine the usefulness of a label refinement for a given event log from a process perspective. Based on data from smart homes, we show how our statistical evaluation method for label refinements can be used in practice. Our method was able to select two label refinements out of a set of candidate label refinements that both had a positive effect on model precision.

Consider a number of workers running SGD independently on the same pool of data and averaging the models every once in a while -- a common but not well understood practice. We study model averaging as a variance-reducing mechanism and describe two ways in which the frequency of averaging affects convergence. For convex objectives, we show the benefit of frequent averaging depends on the gradient variance envelope. For non-convex objectives, we illustrate that this benefit depends on the presence of multiple globally optimal points. We complement our findings with multicore experiments on both synthetic and real data.

We introduce structured prediction energy networks (SPENs), a flexible framework for structured prediction. A deep architecture is used to define an energy function of candidate labels, and then predictions are produced by using back-propagation to iteratively optimize the energy with respect to the labels. This deep architecture captures dependencies between labels that would lead to intractable graphical models, and performs structure learning by automatically learning discriminative features of the structured output. One natural application of our technique is multi-label classification, which traditionally has required strict prior assumptions about the interactions between labels to ensure tractable learning and prediction. We are able to apply SPENs to multi-label problems with substantially larger label sets than previous applications of structured prediction, while modeling high-order interactions using minimal structural assumptions. Overall, deep learning provides remarkable tools for learning features of the inputs to a prediction problem, and this work extends these techniques to learning features of structured outputs. Our experiments provide impressive performance on a variety of benchmark multi-label classification tasks, demonstrate that our technique can be used to provide interpretable structure learning, and illuminate fundamental trade-offs between feed-forward and iterative structured prediction.

In text classification, dictionaries can be used to define human-comprehensible features. We propose an improvement to dictionary features called smoothed dictionary features. These features recognize document contexts instead of n-grams. We describe a principled methodology to solicit dictionary features from a teacher, and present results showing that models built using these human-comprehensible features are competitive with models trained with Bag of Words features.