Playing violent'shooter' video games can damage the brain and may even increase the risk of Alzheimer's disease, brain scans suggest. Researchers at the University of Montreal got around 100 people to play a range of popular'shooter' games such as Call Of Duty, Killzone and Borderlands 2 for a total of 90 hours. They also gave them copes of non-violent games from the Super Mario series. By the end of the study, they found that people who habitually played action games had fewer neurons in their hippocampus, a key memory center in the brain. However, those who played non-violent games ended the study with more essential gray matter in their brains.

The images you put up on Instagram could be used to diagnose if you're depressed. Psychologists say they can contain several red flags. They are darker, more likely to be black and white, and feature fewer people, as sufferers retreat from social contact. A'sadness selfie' may even exist, highlighting that someone is struggling. Researchers have now created a computer to detect depression in photos uploaded to social media - and they say it can correctly pick up on clues 70 per cent of the time.

If you are looking for a flash of inspiration, it might be at the bottom of a pint of beer. A study has found a drink can help us think more creatively, by freeing up the brain to think in a different way. It does not require industrial strength quantities, such as those enjoyed by the great poet Dylan Thomas or novelist Marguerite Duras. But the equivalent of a pint of beer or a small glass of wine was found by Austrian scientists to help people unleash their creative side. The researchers gave 70 people a drink of either beer or non-alcohol beer, which they were unable to distinguish between.

Infants growing up bilingual have the capability to differentiate between the two languages, even when they are as young as 20 months old, a new study published Tuesday by Princeton University found. Researchers at Princeton Baby Lab studied how babies and young children learn to see, talk and comprehend the world. International researchers teamed up with researchers from Princeton University and found infants as young as 20 months of age could accurately and efficiently process two languages separately. "By 20 months, bilingual babies already know something about the differences between words in their two languages," Casey Lew-Williams, an assistant professor of psychology and co-director of the Princeton Baby Lab, said. "They do not think that'dog' and'chien' [French] are just two versions of the same thing," he added.

But you don't know where to start, or perhaps you have read some theory, but don't know how to implement what you have learned. This tutorial will help you break the ice, and walk you through the complete process from importing and analysing a dataset to implementing and training a few different well known classification algorithms and assessing their performance. I'll be using a minimal amount of discrete mathematics, and aim to express details using intuition, and concrete examples instead of dense mathematical formulas. You can read why here. We will be classifying flower-species based on their sepal and petal characteristics using the Iris flower dataset which you can download from Kaggle here. Kaggle, if you haven't heard of it, has a ton of cool open datasets, and is a place where data scientists share their work which can be a valuable resource when learning.

Surrogate models provide a low computational cost alternative to evaluating expensive functions. The construction of accurate surrogate models with large numbers of independent variables is currently prohibitive because it requires a large number of function evaluations. Gradient-enhanced kriging has the potential to reduce the number of function evaluations for the desired accuracy when efficient gradient computation, such as an adjoint method, is available. However, current gradient-enhanced kriging methods do not scale well with the number of sampling points due to the rapid growth in the size of the correlation matrix where new information is added for each sampling point in each direction of the design space. They do not scale well with the number of independent variables either due to the increase in the number of hyperparameters that needs to be estimated. To address this issue, we develop a new gradient-enhanced surrogate model approach that drastically reduced the number of hyperparameters through the use of the partial-least squares method that maintains accuracy. In addition, this method is able to control the size of the correlation matrix by adding only relevant points defined through the information provided by the partial-least squares method. To validate our method, we compare the global accuracy of the proposed method with conventional kriging surrogate models on two analytic functions with up to 100 dimensions, as well as engineering problems of varied complexity with up to 15 dimensions. We show that the proposed method requires fewer sampling points than conventional methods to obtain the desired accuracy, or provides more accuracy for a fixed budget of sampling points. In some cases, we get over 3 times more accurate models than a bench of surrogate models from the literature, and also over 3200 times faster than standard gradient-enhanced kriging models.

The Nystr\"om method is a popular technique for computing fixed-rank approximations of large kernel matrices using a small number of landmark points. In practice, to ensure high quality approximations, the number of landmark points is chosen to be greater than the target rank. However, the standard Nystr\"om method uses a sub-optimal procedure for rank reduction mainly due to its simplicity. In this paper, we highlight the drawbacks of standard Nystr\"om in terms of poor performance and lack of theoretical guarantees. To address these issues, we present an efficient method for generating improved fixed-rank Nystr\"om approximations. Theoretical analysis and numerical experiments are provided to demonstrate the advantages of the modified method over the standard Nystr\"om method. Overall, the aim of this paper is to convince researchers to use the modified method, as it has nearly identical computational complexity, is easy to code, and has greatly improved accuracy in many cases.

Sensing in complex systems requires large-scale information exchange and on-the-go communications over heterogeneous networks and integrated processing platforms. Many networked cyber-physical systems exhibit hierarchical infrastructures of information flows, which naturally leads to a multi-level tree-like information structure in which each level corresponds to a particular scale of representation. This work focuses on the multiscale fusion of data collected at multiple levels of the system. We propose a multiscale state-space model to represent multi-resolution data over the hierarchical information system and formulate a multi-stage dynamic zero-sum game to design a multi-scale $H_{\infty}$ robust filter. We present numerical experiments for one and two-dimensional signals and provide a comparative analysis of the minimax filter with the standard Kalman filter to show the improvement in signal-to-noise ratio (SNR).

We approach structured output prediction by optimizing a deep value network (DVN) to precisely estimate the task loss on different output configurations for a given input. Once the model is trained, we perform inference by gradient descent on the continuous relaxations of the output variables to find outputs with promising scores from the value network. When applied to image segmentation, the value network takes an image and a segmentation mask as inputs and predicts a scalar estimating the intersection over union between the input and ground truth masks. For multi-label classification, the DVN's objective is to correctly predict the F1 score for any potential label configuration. The DVN framework achieves the state-of-the-art results on multi-label prediction and image segmentation benchmarks.

We consider the demixing problem of two (or more) structured high-dimensional vectors from a limited number of nonlinear observations where this nonlinearity is due to either a periodic or an aperiodic function. We study certain families of structured superposition models, and propose a method which provably recovers the components given (nearly) $m = \mathcal{O}(s)$ samples where $s$ denotes the sparsity level of the underlying components. This strictly improves upon previous nonlinear demixing techniques and asymptotically matches the best possible sample complexity. We also provide a range of simulations to illustrate the performance of the proposed algorithms.