Audiobooks are more emotionally engaging than TV and film – even if you don't realise it, according to a landmark new study. The new research from UCL suggests that having a book read to you causes physiological changes including an increased heart rate and heat spreading through your body. During the experiment, scientists had 103 participants of various ages listen to a range of different books, and compared their responses to how they felt when they watched the same scene in a film or TV adaptation. The study included emotional scenes from Game of Thrones and the Girl on the Train, for instance, both from the original book and their hugely popular adaptations. The I.F.O. is fuelled by eight electric engines, which is able to push the flying object to an estimated top speed of about 120mph.

Convolutional neural networks (CNN) have been shown to provide a good solution for classification problems that utilize data obtained from vibrational spectroscopy. Moreover, CNNs are capable of identification from noisy spectra without the need for additional preprocessing. However, their application in practical spectroscopy is limited due to two shortcomings. The effectiveness of the classification using CNNs drops rapidly when only a small number of spectra per substance are available for training (which is a typical situation in real applications). Additionally, to accommodate new, previously unseen substance classes, the network must be retrained which is computationally intensive. Here we address these issues by reformulating a multi-class classification problem with a large number of classes, but a small number of samples per class, to a binary classification problem with sufficient data available for representation learning. Namely, we define the learning task as identifying pairs of inputs as belonging to the same or different classes. We achieve this using a Siamese convolutional neural network. A novel sampling strategy is proposed to address the imbalance problem in training the Siamese Network. The trained network can effectively classify samples of unseen substance classes using just a single reference sample (termed as one-shot learning in the machine learning community). Our results demonstrate better accuracy than other practical systems to date, while allowing effortless updates of the system's database with novel substance classes.

Manifold learning offers nonlinear dimensionality reduction of high-dimensional datasets. In this paper, we bring geometry processing to bear on manifold learning by introducing a new approach based on metric connection for generating a quasi-isometric, low-dimensional mapping from a sparse and irregular sampling of an arbitrary manifold embedded in a high-dimensional space. Geodesic distances of discrete paths on the input pointset are evaluated through "parallel transport unfolding" (PTU) to offer robustness to poor sampling and arbitrary topology. Our new geometric procedure exhibits the same strong resilience to noise as one of the staples of manifold learning, the Isomap algorithm, as it also exploits all pairwise geodesic distances to compute a low-dimensional embedding. While Isomap is limited to geodesically-convex sampled domains, parallel transport unfolding does not suffer from this crippling limitation, resulting in an improved robustness to irregularity and voids in the sampling. Moreover, it involves only simple linear algebra, significantly improves the accuracy of all pairwise geodesic distance approximations, and has the same computational complexity as Isomap. Finally, we show that our connection-based distance estimation can be used for faster variants of Isomap such as L-Isomap.

Sliced inverse regression (SIR) is a pioneer tool for supervised dimension reduction. It identifies the effective dimension reduction space, the subspace of significant factors with intrinsic lower dimensionality. In this paper, we propose to refine the SIR algorithm through an overlapping slicing scheme. The new algorithm, called overlapping sliced inverse regression (OSIR), is able to estimate the effective dimension reduction space and determine the number of effective factors more accurately. We show that such overlapping procedure has the potential to identify the information contained in the derivatives of the inverse regression curve, which helps to explain the superiority of OSIR. We also prove that OSIR algorithm is $\sqrt n $-consistent and verify its effectiveness by simulations and real applications.

Distributed model training suffers from communication overheads due to frequent gradient updates transmitted between compute nodes. To mitigate these overheads, several studies propose the use of sparsified stochastic gradients. We argue that these are facets of a general sparsification method that can operate on any possible atomic decomposition. Notable examples include element-wise, singular value, and Fourier decompositions. We present ATOMO, a general framework for atomic sparsification of stochastic gradients. Given a gradient, an atomic decomposition, and a sparsity budget, ATOMO gives a random unbiased sparsification of the atoms minimizing variance. We show that methods such as QSGD and TernGrad are special cases of ATOMO and show that sparsifiying gradients in their singular value decomposition (SVD), rather than the coordinate-wise one, can lead to significantly faster distributed training.

Recent years have witnessed the unprecedented rising of time series from almost all kindes of academic and industrial fields. Various types of deep neural network models have been introduced to time series analysis, but the important frequency information is yet lack of effective modeling. In light of this, in this paper we propose a wavelet-based neural network structure called multilevel Wavelet Decomposition Network (mWDN) for building frequency-aware deep learning models for time series analysis. mWDN preserves the advantage of multilevel discrete wavelet decomposition in frequency learning while enables the fine-tuning of all parameters under a deep neural network framework. Based on mWDN, we further propose two deep learning models called Residual Classification Flow (RCF) and multi-frequecy Long Short-Term Memory (mLSTM) for time series classification and forecasting, respectively. The two models take all or partial mWDN decomposed sub-series in different frequencies as input, and resort to the back propagation algorithm to learn all the parameters globally, which enables seamless embedding of wavelet-based frequency analysis into deep learning frameworks. Extensive experiments on 40 UCR datasets and a real-world user volume dataset demonstrate the excellent performance of our time series models based on mWDN. In particular, we propose an importance analysis method to mWDN based models, which successfully identifies those time-series elements and mWDN layers that are crucially important to time series analysis. This indeed indicates the interpretability advantage of mWDN, and can be viewed as an indepth exploration to interpretable deep learning.

Evaluating on adversarial examples has become a standard procedure to measure robustness of deep learning models. Due to the difficulty of creating white-box adversarial examples for discrete text input, most analyses of the robustness of NLP models have been done through black-box adversarial examples. We investigate adversarial examples for character-level neural machine translation (NMT), and contrast black-box adversaries with a novel white-box adversary, which employs differentiable string-edit operations to rank adversarial changes. We propose two novel types of attacks which aim to remove or change a word in a translation, rather than simply break the NMT. We demonstrate that white-box adversarial examples are significantly stronger than their black-box counterparts in different attack scenarios, which show more serious vulnerabilities than previously known. In addition, after performing adversarial training, which takes only 3 times longer than regular training, we can improve the model's robustness significantly.

To be informative, an evaluation must measure how well systems generalize to realistic unseen data. We identify limitations of and propose improvements to current evaluations of text-to-SQL systems. First, we compare human-generated and automatically generated questions, characterizing properties of queries necessary for real-world applications. To facilitate evaluation on multiple datasets, we release standardized and improved versions of seven existing datasets and one new text-to-SQL dataset. Second, we show that the current division of data into training and test sets measures robustness to variations in the way questions are asked, but only partially tests how well systems generalize to new queries; therefore, we propose a complementary dataset split for evaluation of future work. Finally, we demonstrate how the common practice of anonymizing variables during evaluation removes an important challenge of the task. Our observations highlight key difficulties, and our methodology enables effective measurement of future development.

Training large neural networks requires distributing learning across multiple workers, where the cost of communicating gradients can be a significant bottleneck. signSGD alleviates this problem by transmitting just the sign of each minibatch stochastic gradient. We prove that it can get the best of both worlds: compressed gradients and SGD-level convergence rate. The relative $\ell_1/\ell_2$ geometry of gradients, noise and curvature informs whether signSGD or SGD is theoretically better suited to a particular problem. On the practical side we find that the momentum counterpart of signSGD is able to match the accuracy and convergence speed of Adam on deep Imagenet models. We extend our theory to the distributed setting, where the parameter server uses majority vote to aggregate gradient signs from each worker enabling 1-bit compression of worker-server communication in both directions. Using a theorem by Gauss we prove that majority vote can achieve the same reduction in variance as full precision distributed SGD. Thus, there is great promise for sign-based optimisation schemes to achieve fast communication and fast convergence. Code to reproduce experiments is to be found at https://github.com/jxbz/signSGD.

A commonly used strategy for improving optimization algorithms is to restart the algorithm when it is believed to be trapped in an inferior part of the search space. Building on the recent success of Bet-and-Run approaches for restarted local search solvers, we introduce an improved generic Bet-and-Run strategy. The goal is to obtain the best possible results within a given time budget t using a given black-box optimization algorithm. If no prior knowledge about problem features and algorithm behavior is available, the question about how to use the time budget most efficiently arises. We propose to first start k>=1 independent runs of the algorithm during an initialization budget t1=0 time units in doing so), and then continuing these runs for the remaining t3=t-t1-t2 time units. In previous Bet-and-Run strategies, the decision maker D=currentBest would simply select the run with the best- so-far results at negligible time. We propose using more advanced methods to discriminate between "good" and "bad" sample runs, with the goal of increasing the correlation of the chosen run with the a-posteriori best one. We test several different approaches, including neural networks trained or polynomials fitted on the current trace of the algorithm to predict which run may yield the best results if granted the remaining budget. We show with extensive experiments that this approach can yield better results than the previous methods, but also find that the currentBest method is a very reliable and robust baseline approach.