Dynamic functional connectivity (FC) has in recent years become a topic of interest in the neuroimaging community. Several models and methods exist for both functional magnetic resonance imaging (fMRI) and electroencephalography (EEG), and the results point towards the conclusion that FC exhibits dynamic changes. The existing approaches modeling dynamic connectivity have primarily been based on time-windowing the data and k-means clustering. We propose a non-parametric generative model for dynamic FC in fMRI that does not rely on specifying window lengths and number of dynamic states. Rooted in Bayesian statistical modeling we use the predictive likelihood to investigate if the model can discriminate between a motor task and rest both within and across subjects. We further investigate what drives dynamic states using the model on the entire data collated across subjects and task/rest. We find that the number of states extracted are driven by subject variability and preprocessing differences while the individual states are almost purely defined by either task or rest. This questions how we in general interpret dynamic FC and points to the need for more research on what drives dynamic FC.

The counting grid is a grid of microtopics, sparse word/feature distributions. The generative model associated with the grid does not use these microtopics individually, but in predefined groups which can only be (ad)mixed as such. Each allowed group corresponds to one of all possible overlapping rectangular windows into the grid. The capacity of the model is controlled by the ratio of the grid size and the window size. This paper builds upon the basic counting grid model and it shows that hierarchical reasoning helps avoid bad local minima, produces better classification accuracy and, most interestingly, allows for extraction of large numbers of coherent microtopics even from small datasets. We evaluate this in terms of consistency, diversity and clarity of the indexed content, as well as in a user study on word intrusion tasks. We demonstrate that these models work well as a technique for embedding raw images and discuss interesting parallels between hierarchical CG models and other deep architectures.

Bayesian inference has great promise for the privacy-preserving analysis of sensitive data, as posterior sampling automatically preserves differential privacy, an algorithmic notion of data privacy, under certain conditions (Dimitrakakis et al., 2014; Wang et al., 2015b). While this one posterior sample (OPS) approach elegantly provides privacy "for free," it is data inefficient in the sense of asymptotic relative efficiency (ARE). We show that a simple alternative based on the Laplace mechanism, the workhorse of differential privacy, is as asymptotically efficient as non-private posterior inference, under general assumptions. This technique also has practical advantages including efficient use of the privacy budget for MCMC. We demonstrate the practicality of our approach on a time-series analysis of sensitive military records from the Afghanistan and Iraq wars disclosed by the Wikileaks organization.

We address an important problem in sequence-to-sequence (Seq2Seq) learning referred to as copying, in which certain segments in the input sequence are selectively replicated in the output sequence. A similar phenomenon is observable in human language communication. For example, humans tend to repeat entity names or even long phrases in conversation. The challenge with regard to copying in Seq2Seq is that new machinery is needed to decide when to perform the operation. In this paper, we incorporate copying into neural network-based Seq2Seq learning and propose a new model called CopyNet with encoder-decoder structure. CopyNet can nicely integrate the regular way of word generation in the decoder with the new copying mechanism which can choose sub-sequences in the input sequence and put them at proper places in the output sequence. Our empirical study on both synthetic data sets and real world data sets demonstrates the efficacy of CopyNet. For example, CopyNet can outperform regular RNN-based model with remarkable margins on text summarization tasks.

Footage of Bryde's whales feeding has been caught on camera. It was filmed with the use of a drone in research that paves the way for further studies in marine animal behaviour. Dr Barbara Bollard-Breen a senior lecturer at Auckland University of Technology, spoke to BBC News about the importance of the footage.

Google DeepMind has poured cold water on rumours its AlphaGo Artificial Intelligence Go-playing program will face off against Chinese Go champion Ke Jie. "Contrary to internet rumours, we've not decided yet what to do next with #AlphaGo, once we have, there will be an official announcement here," Hassabis tweeted on Monday. The news was originally released by Yang Junan, secretary general of the International Go Federation, at a news conference for the 37th World Amateur Go Championship, on June 4. Yang said representatives had been in contact with the team behind AlphaGo and would set up a match by the end of this year. AlphaGo defeated South Korean Go grandmaster Lee Sedol 4-1 at the Google DeepMind Challenge Match held in March, sparking global interest in AI.

Given the fact that a universal income – at anything but the stingiest level – would involve redistributing untold billions in taxation from rich to poor, this seems less like "socialism with an iPad", in McDonnell's phrase, and more like socialism plain and simple. Which is perhaps why Swiss voters gave the idea a gigantic raspberry in a referendum at the weekend. Experts are increasingly convinced that computers will soon be able to do many human jobs in less time and for less money – in other words, that the pace of technological change will outpace the labour market's ability to cope. The same factories in China that undercut the West with their low labour costs are already replacing those same workers with even cheaper robots; soon, middle-class professionals in Brighton or Boston could find that they are just as vulnerable to being given the P45 by their PC.

Don't be surprised if you see a very large, very unusual drone flying through Nevada's skies. The state's Institute for Autonomous Systems has given China's EHang permission to test fly its passenger-toting 184 drone later this year. In addition to providing basic clearance, the move will also have the Institute create criteria that shows the airworthiness of the autonomous single-seater to the Federal Aviation Administration. It's not certain just where the 184 will fly, although it'll sometimes need restricted airspace. EHang won't just be flying in the empty desert, then.

Travis Kalanick, Founder and CEO of Uber, delivers a speech at the Institute of Directors Convention at the Royal Albert Hall, Central London, Britain, 03 October 2014. Uber may be increasing its testing of self-driving cars, but it still has plenty of work to do to keep its human drivers happy. The ride-sharing leader announced a host of new initiatives that are sure to please drivers, giving them more control over the rides they drive. The moves, which the company started rolling out this week, run the gamut in improving the Uber driving experience. One new feature allows Uber drivers to pause new requests from coming in while they are finishing a route, while a second lets drivers set a "driver destination," or letting a driver twice a day input where they want to go and matching them only with riders on that route.

At least for the 3,000 software engineers whose tasks are about to be replaced by an artificial intelligence (AI) tool at Wipro. Mint's Varun Sood reports that Wipro is all set to use "Holmes," its artificial intelligence wizard to automate certain projects and this, in turn, will free up 3,000 engineers from "mundane" software maintenance jobs. This step is expected to save the company about 46.5 million as it tries to strengthen profitability in what is becoming an increasingly uncertain environment for outsourcing companies. Wipro also plans to sell this artificial intelligence tool to new and existing clients this year and earn anywhere between 60 million to 70 million, according to the report. The AI platform will be deployed at the company's'fixed-price' projects, which employ about 30,000 people.