Gradient boosting of regression trees is a competitive procedure for learning predictive models of continuous data that fits the data with an additive non-parametric model. The classic version of gradient boosting assumes that the data is independent and identically distributed. However, relational data with interdependent, linked instances is now common and the dependencies in such data can be exploited to improve predictive performance. Collective inference is one approach to exploit relational correlation patterns and significantly reduce classification error. However, much of the work on collective learning and inference has focused on discrete prediction tasks rather than continuous. %target values has not got that attention in terms of collective inference. In this work, we investigate how to combine these two paradigms together to improve regression in relational domains. Specifically, we propose a boosting algorithm for learning a collective inference model that predicts a continuous target variable. In the algorithm, we learn a basic relational model, collectively infer the target values, and then iteratively learn relational models to predict the residuals. We evaluate our proposed algorithm on a real network dataset and show that it outperforms alternative boosting methods. However, our investigation also revealed that the relational features interact together to produce better predictions.

We derive a new discrepancy statistic for measuring differences between two probability distributions based on combining Stein's identity with the reproducing kernel Hilbert space theory. We apply our result to test how well a probabilistic model fits a set of observations, and derive a new class of powerful goodness-of-fit tests that are widely applicable for complex and high dimensional distributions, even for those with computationally intractable normalization constants. Both theoretical and empirical properties of our methods are studied thoroughly.

Cascading chains of events are a salient feature of many real-world social, biological, and financial networks. In social networks, social reciprocity accounts for retaliations in gang interactions, proxy wars in nation-state conflicts, or Internet memes shared via social media. Neuron spikes stimulate or inhibit spike activity in other neurons. Stock market shocks can trigger a contagion of volatility throughout a financial network. In these and other examples, only individual events associated with network nodes are observed, usually without knowledge of the underlying dynamic relationships between nodes. This paper addresses the challenge of tracking how events within such networks stimulate or influence future events. The proposed approach is an online learning framework well-suited to streaming data, using a multivariate Hawkes point process model to encapsulate autoregressive features of observed events within the social network. Recent work on online learning in dynamic environments is leveraged not only to exploit the dynamics within the underlying network, but also to track the network structure as it evolves. Regret bounds and experimental results demonstrate that the proposed method performs nearly as well as an oracle or batch algorithm.

Most Machine Learning (ML) methods, from clustering to classification, rely on a distance function to describe relationships between datapoints. For complex datasets it is hard to avoid making some arbitrary choices when defining a distance function. To compare images, one must choose a spatial scale, for signals, a temporal scale. The right scale is hard to pin down and it is preferable when results do not depend too tightly on the exact value one picked. Topological data analysis seeks to address this issue by focusing on the notion of neighbourhood instead of distance. It is shown that in some cases a simpler solution is available. It can be checked how strongly distance relationships depend on a hyperparameter using dimensionality reduction. A variant of dynamical multi-dimensional scaling (MDS) is formulated, which embeds datapoints as curves. The resulting algorithm is based on the Concave-Convex Procedure (CCCP) and provides a simple and efficient way of visualizing changes and invariances in distance patterns as a hyperparameter is varied. A variant to analyze the dependence on multiple hyperparameters is also presented. A cMDS algorithm that is straightforward to implement, use and extend is provided. To illustrate the possibilities of cMDS, cMDS is applied to several real-world data sets.

The National Highway Transportation Safety Board is opening a preliminary evaluation into Tesla's autopilot feature, after the fatal crash of a Model S that was in self-driving mode, the electric automaker said Thursday. According to a blog post from Tesla Motors Inc., the car was on a unnamed, divided highway when a tractor trailer drove across the road perpendicular to the Model S. "Neither Autopilot nor the driver noticed the white side of the tractor trailer against a brightly lit sky, so the brake was not applied," Tesla said in the post. The Model S passed under the trailer, with the bottom of the trailer impacting the windshield of the Model S, Tesla said. Tesla said this was the first fatality in which the autopilot feature was activated, with more than 130 million miles driven using that feature. The Palo Alto automaker said it informed NHTSA about the incident "immediately after it occurred."

Watchwith has acquired the Arris Media Analysis Framework (MAF), a cloud-based machine learning and automated metadata generation platform. MAF was developed in Arris research labs, and the technology analyzes, tags and describes video at a frame level, which eliminates manual tagging. The companies have integrated the automation technology into Watchwith's data-driven advanced advertising products. "What used to potentially require thousands of man-hours is now an automated process within the Watchwith platform," Watchwith says in a statement. The combined solution is able to automatically determine the optimal timing and location within a TV episode to deliver in-program advertising and tune-in messages.

Chemists and mathematicians from the Skolkovo Institute of Science and Technology (Skoltech) and Moscow State Universite (MSU) have suggested checking the composition of medical plants by means of machine learning technologies, the Skoltech press service said. They have come up with automatizing computer assisted data analysis based on high-performance liquid chromatography and mass spectrometry. "Machine learning is when a computer can be taught to analyze the chemical composition of herbal medicine based on the previously known data on chemical analysis," Skoltech said. According to the researchers, the market of herbal remedies has been rapidly developing in the recent years, as it provides an alternative to synthetic drugs. But there are still no existing effective methods of plant material quality control.

The company reported on the incident in a blog post. On May 7, Ohio resident Joshua Brown, 45, was in the driver's seat of a Tesla Model S in Williston, Florida. The car's Autopilot was engaged when a tractor-trailer made a left turn in front of the electric vehicle. Brown was killed "when he drove under the trailer," the Levy Journal reported at the time."The After striking the underside of the trailer, the car then continued driving until it left the road, struck a fence, smashed through two other fences and struck a power pole. On Thursday, Tesla announced that the NHTSA had opened a preliminary evaluation on Wednesday into the performance of Autopilot in the crash. Following the company's standard practice, it had informed the federal agency of the accident immediately after it occurred. "Neither Autopilot nor the driver noticed the white side of the tractor trailer against a brightly lit sky, so the brake was not applied.

In addition, the built-in artificial intelligence will be able to predict situations, producing ready-made proposals for the border protection. "The system is fully based on domestic policy decisions that ensure protection of information resources against data loss, hackers and other unauthorized interventions," the press service quoted the deputy director of OPK Sergei Skokov as saying. The developers also noted that the new system is intended not only to collect different types of information, but also contains elements of artificial intelligence which will allow for analysis and forecasting of the situation and work out proposals for the protection of borders, by calculating steps and routes that offenders may take, as well as the necessary measures to prevent malicious acts, including the assessment of possible risks. The state borders need protection due to ever rising threats. Since the beginning of this year in the Rostov region, more than 60 "wanted" persons were found and arrested.

Stephen Hawking made headlines at the end of 2014 when, in a BBC interview, he said that we should be very wary of developing "full artificial intelligence" as it "could spell the end of the human race." His doomsday musings were hardly original. SpaceX's Elon Musk had said the same thing earlier that year, warning that AI is "potentially more dangerous than nukes." The worrisome idea of computers possessing greater than human intelligence, coupled with a sudden independent consciousness, was first termed "The Singularity" back in 1993, in a paper by the computer scientist Vernor Vinge. And while his initial predictions about vast computer improvements merely mirrored the foresight of others – like the expected frequent doubling in computer power envisioned by Intel co-founder Gordon Moore in 1965 – Vinge believed it would lead to "change comparable to the rise of human life on Earth." As we all know, computers already control and facilitate much of our daily life from banking to robotic automobile assembly, and no one wants to return to the old days of manual drudgery for menial tasks like repetitive spot welding.