Understanding human driving behavior is important for autonomous vehicles. In this paper, we propose an interpretable human behavior model in interactive driving scenarios based on the cumulative prospect theory (CPT). As a non-expected utility theory, CPT can well explain some systematically biased or ``irrational'' behavior/decisions of human that cannot be explained by the expected utility theory. Hence, the goal of this work is to formulate the human drivers' behavior generation model with CPT so that some ``irrational'' behavior or decisions of human can be better captured and predicted. Towards such a goal, we first develop a CPT-driven decision-making model focusing on driving scenarios with two interacting agents. A hierarchical learning algorithm is proposed afterward to learn the utility function, the value function, and the decision weighting function in the CPT model. A case study for roundabout merging is also provided as verification. With real driving data, the prediction performances of three different models are compared: a predefined model based on time-to-collision (TTC), a learning-based model based on neural networks, and the proposed CPT-based model. The results show that the proposed model outperforms the TTC model and achieves similar performance as the learning-based model with much less training data and better interpretability.

Non-local self-similarity is well-known to be an effective prior for the image denoising problem. However, little work has been done to incorporate it in convolutional neural networks, which surpass non-local model-based methods despite only exploiting local information. In this paper, we propose a novel end-to-end trainable neural network architecture employing layers based on graph convolution operations, thereby creating neurons with non-local receptive fields. The graph convolution operation generalizes the classic convolution to arbitrary graphs. In this work, the graph is dynamically computed from similarities among the hidden features of the network, so that the powerful representation learning capabilities of the network are exploited to uncover self-similar patterns. We introduce a lightweight Edge-Conditioned Convolution which addresses vanishing gradient and over-parameterization issues of this particular graph convolution. Extensive experiments show state-of-the-art performance with improved qualitative and quantitative results on both synthetic Gaussian noise and real noise.

In the last few years, neural networks have been intensively used to develop meaningful distributed representations of words and contexts around them. When these representations, also known as "embed-dings", are learned from unsupervised large corpora, they can be transferred to different tasks with positive effects in terms of performances, especially when only a few supervisions are available. In this work, we further extend this concept, and we present an unsupervised neural architecture that jointly learns word and context embeddings, processing words as sequences of characters. This allows our model to spot the regularities that are due to the word morphology, and to avoid the need of a fixed-sized input vocabulary of words. We show that we can learn compact encoders that, despite the relatively small number of parameters, reach high-level performances in downstream tasks, comparing them with related state-of-the-art approaches or with fully supervised methods. Keywords: Recurrent Neural Networks, Unsupervised Learning, Word and Context Embeddings, Natural Language Processing, Deep Learning 1 Introduction Recent advances in Natural Language Processing (NLP) are characterized by the development of techniques that compute powerful word embeddings and by the extensive use of neural language models. Word Embeddings (WEs) aim at representing individual words in a low-dimensional continuous space, in order to exploit its topological properties to model semantic or grammatical relationships between different words. In particular, they are based on the assumption that functionally or semantically related words appear in similar contexts. Despite the idea of continuous word representations was proposed a several years ago [4], their importance became strongly popular mostly after the work ofnull This is a post-peer-review, pre-copyedit version of an article published in LNCS, volume 11141.

Abstract: The shortage o f high - resolution urban digital elevation model ( DEM) datasets has been a challenge for modelling urban flood and managing its risk . A solution is to develop effective approaches to r econstruct high - resolution DEMs from their low - resolution equivalents that are more widely available . However, the current high - resolution DEM reconstruction approaches mainly focus on natural topography . F ew attempts have been made for urban topography which is typically a n integration of complex man - made and natural features . T his study proposes a novel multi - scale mapping approach based on convolutional neural network (CNN) to deal with the complex characteristics of urban topography and reconstruct high - resolution urban DEMs . T he proposed multi - scale CNN model is first ly trained using urban DEMs that contain topographic features at different resolutions, and then used to reconstruct the urban DEM at a specified (high) resolution from a low - resolution equivalent . A two - level accuracy assessment approach is also designed to evaluate the performance of the proposed urban DEM reconstruction method, in terms of numerical accuracy and morphological accuracy . Compared with other commonly used met hods, the current CNN based approach produces superior results, provid ing a cost - effective innovative method to acquire high - resolution DEMs in other data - scarce environment s . Introduction Digital elevation model s (DEM s) have been widely used in many fields such as l andform evolution, soil erosion modeling and other geo - simulation s ( Bishop et al., 2012; Liu et al., 2015; Mondal et al., 2017; Li and Wong, 2010) . In p articular, DEMs provide indispensable data to support water resources management and flood risk assessment (Moore et al., 1991; O'Loughlin et al., 2016) . I n urban flood risk assessment, the availability of high - resolution urban DEMs is crucial for the accurate representation of complex urban topographic features and required for a reliable prediction of flood inundation to inform risk calculation ( Ramirez et al., 2016; Leitão and de Sousa, 2018) .

The vast majority of successful deep neural networks are trained using variants of stochastic gradient descent (SGD) algorithms. Recent attempts to improve SGD can be broadly categorized into two approaches: (1) adaptive learning rate schemes, such as AdaGrad and Adam, and (2) accelerated schemes, such as heavy-ball and Nesterov momentum. In this paper, we propose a new optimization algorithm, Lookahead, that is orthogonal to these previous approaches and iteratively updates two sets of weights. Intuitively, the algorithm chooses a search direction by \emph{looking ahead} at the sequence of "fast weights" generated by another optimizer. We show that Lookahead improves the learning stability and lowers the variance of its inner optimizer with negligible computation and memory cost. We empirically demonstrate Lookahead can significantly improve the performance of SGD and Adam, even with their default hyperparameter settings on ImageNet, CIFAR-10/100, neural machine translation, and Penn Treebank.

Mode decomposition is a prototypical pattern recognition problem that can be addressed from the (a priori distinct) perspectives of numerical approximation, statistical inference and deep learning. Could its analysis through these combined perspectives be used as a Rosetta stone for deciphering mechanisms at play in deep learning? Motivated by this question we introduce programmable and interpretable regression networks for pattern recognition and address mode decomposition as a prototypical problem. The programming of these networks is achieved by assembling elementary modules decomposing and recomposing kernels and data. These elementary steps are repeated across levels of abstraction and interpreted from the equivalent perspectives of optimal recovery, game theory and Gaussian process regression (GPR). The prototypical mode/kernel decomposition module produces an optimal approximation $(w_1,w_2,\cdots,w_m)$ of an element $(v_1,v_2,\ldots,v_m)$ of a product of Hilbert subspaces of a common Hilbert space from the observation of the sum $v:=v_1+\cdots+v_m$. The prototypical mode/kernel recomposition module performs partial sums of the recovered modes $w_i$ based on the alignment between each recovered mode $w_i$ and the data $v$. We illustrate the proposed framework by programming regression networks approximating the modes $v_i= a_i(t)y_i\big(\theta_i(t)\big)$ of a (possibly noisy) signal $\sum_i v_i$ when the amplitudes $a_i$, instantaneous phases $\theta_i$ and periodic waveforms $y_i$ may all be unknown and show near machine precision recovery under regularity and separation assumptions on the instantaneous amplitudes $a_i$ and frequencies $\dot{\theta}_i$. The structure of some of these networks share intriguing similarities with convolutional neural networks while being interpretable, programmable and amenable to theoretical analysis.

We propose a metric, Layer Saturation, defined as the proportion of the number of eigenvalues needed to explain 99% of the variance of the latent representations, for analyzing the learned representations of neural network layers. Saturation is based on spectral analysis and can be computed efficiently, making live analysis of the representations practical during training. We provide an outlook for future applications of this metric by outlining the behaviour of layer saturation in different neural architectures and problems. We further show that saturation is related to the generalization and predictive performance of neural networks.

Multimorbidity, or the presence of several medical conditions in the same individual, have been increasing in the population both in absolute and relative terms. However, multimorbidity remains poorly understood, and the evidence from existing research to describe its burden, determinants and consequences have been limited. Many of these studies are often cross-sectional and do not explicitly account for multimorbidity patterns' evolution over time. Some studies were based on small datasets, used arbitrary or narrow age range, or lacked appropriate clinical validations. In this study, we applied Non-negative Matrix Factorisation (NMF) in a novel way to one of the largest electronic health records (EHR) databases in the world (with 4 million patients), for simultaneously modelling disease clusters and their role in one's multimorbidity over time. Furthermore, we demonstrated how the temporal characteristics that our model associates with each disease cluster can help mine disease trajectories/networks and generate new hypotheses for the formation of multimorbidity clusters as a function of time/ageing. Our results suggest that our method's ability to learn the underlying dynamics of diseases can provide the field with a novel data-driven / exploratory way of learning the patterns of multimorbidity and their interactions over time.

Improving generalization is one of the main challenges for training deep neural networks on classification tasks. In particular, a number of techniques have been proposed, aiming to boost the performance on unseen data: from standard data augmentation techniques to the $\ell_2$ regularization, dropout, batch normalization, entropy-driven SGD and many more.\\ In this work we propose an elegant, simple and principled approach: post-synaptic potential regularization (PSP). We tested this regularization on a number of different state-of-the-art scenarios. Empirical results show that PSP achieves a classification error comparable to more sophisticated learning strategies in the MNIST scenario, while improves the generalization compared to $\ell_2$ regularization in deep architectures trained on CIFAR-10.

We designed and implemented a CUDA port of the Atari Learning Environment (ALE), a system for developing and evaluating deep reinforcement algorithms using Atari games. Our CUDA Learning Environment (CuLE) overcomes many limitations of existing CPUbased Atari emulators and scales naturally to multi-GPU systems. It leverages the parallelization capability of GPUs to run thousands of Atari games simultaneously; by rendering frames directly on the GPU, CuLE avoids the bottleneck arising from the limited CPU-GPU communication bandwidth. Figure 1: In a typical DRL system, environments run As a result, CuLE is able to generate between 40M on CPUs, whereas GPUs execute DNN operations. The and 190M frames per hour using a single GPU, a finding limited CPU-GPU communication bandwidth and small that could be previously achieved only through a cluster set of CPU environments prevent full GPU utilization. of CPUs. We demonstrate the advantages of CuLE by effectively training agents with traditional deep reinforcement learning algorithms and measuring the utilization benchmark for DRL [4, 14], and still represent a challenging and throughput of the GPU. Our analysis further highlights set for the development of new DRL methods.