The tech giant's AI lab DeepMind has developed a new tool which flagged fewer false positives & negatives in a trial study

In this research, an extensive literature review was performed on the recent developments of the ambulance routing problem (ARP) and ambulance location problem (ALP). Both are respective modifications of the vehicle routing problem (VRP) and maximum covering problem (MCP), with modifications to objective functions and constraints. Although alike, a key distinction is emergency service systems (EMS) are considered critical and the optimization of these has become all the more important as a result. Similar to their parent problems, these are NP-hard and must resort to approximations if the space size is too large. Much of the current work has simply been on modifying existing systems through simulation to achieve a more acceptable result. There has been attempts towards using meta-heuristics, though practical experimentation is lacking when compared to VRP or MCP. The contributions of this work are a comprehensive survey of current methodologies, summarized models, and suggested future improvements.

Graph neural networks (GNN) rely on graph operations that include neural network training for various graph related tasks. Recently, several attempts have been made to apply the GNNs to functional magnetic resonance image (fMRI) data. Despite the recent progress, a common limitation is its difficulty to explain the classification results in a neuroscientifically explainable way. Here, we develop a framework for analyzing the fMRI data using the Graph Isomorphism Network (GIN), which was recently proposed as a state-of-the-art GNN for graph classification. One important observation in this paper is that the GIN is a realization of convolutional neural network (CNN) with two-tab filters in the graph space where the shift operation is realized using the adjacent matrix. Based on this observation, we visualize the important regions of the brain by a saliency mapping method of the trained GIN. We validate our proposed framework using large-scale resting-state fMRI data for classifying the sex of the subject based on the graph structure of the brain. The experiment was consistent with our expectation such that the obtained saliency map show high correspondence with previous neuroimaging evidences related to sex differences.

Spin-glasses are universal models that can capture complex behavior of many-body systems at the interface of statistical physics and computer science including discrete optimization, inference in graphical models, and automated reasoning. Computing the underlying structure and dynamics of such complex systems is extremely difficult due to the combinatorial explosion of their state space. Here, we develop deep generative continuous spin-glass distributions with normalizing flows to model correlations in generic discrete problems. We use a self-supervised learning paradigm by automatically generating the data from the spin-glass itself. We demonstrate that key physical and computational properties of the spin-glass phase can be successfully learned, including multi-modal steady-state distributions and topological structures among metastable states. Remarkably, we observe that the learning itself corresponds to a spin-glass phase transition within the layers of the trained normalizing flows. The inverse normalizing flows learns to perform reversible multi-scale coarse-graining operations which are very different from the typical irreversible renormalization group techniques.

Time series forecasting admits a wide range of applications from signal processing, pattern recognition and weather forecasting to mathematical finance, to name only a few. Machine learning techniques for time-series forecasting have been widely studied [1, 2]. The traditional recurrent approaches towards sequence modeling tasks [1, 2] have been recently challenged by convolutional network architectures [3-6]. Latter compete in the categories speed and precision and regularly outperform conventional recurrent approaches such as LSTM's, GRU's or RNN's [7-12]. In particular, those convolutional architectures may overcome the deficiencies of recurrent networks to handle long and multi-scale sequences with increased receptive fields [3, 13, 14]. For time sequences of images convolutional LSTM's aim to combine the best of both worlds [15, 16]. In this work we present a novel approach to utilize convolutional neural networks for image sequence as well as general sequence forecasting tasks. In contrast to the recent serge in causal "dilated" convolutional networks [3-6, 13, 14, 17, 18] our approach is closer in spirit to non-casual architectures [19-22]. However, our architecture distinguishes itself by its composite design for time series forecasting, see fig.

The gradient-based optimization method for deep machine learning models suffers from gradient vanishing and exploding problems, particularly when the computational graph becomes deep. In this work, we propose the tangent-space gradient optimization (TSGO) for the probabilistic models to keep the gradients from vanishing or exploding. The central idea is to guarantee the orthogonality between the variational parameters and the gradients. The optimization is then implemented by rotating parameter vector towards the direction of gradient. We explain and testify TSGO in tensor network (TN) machine learning, where the TN describes the joint probability distribution as a normalized state $\left| \psi \right\rangle $ in Hilbert space. We show that the gradient can be restricted in the tangent space of $\left\langle \psi \right.\left| \psi \right\rangle = 1$ hyper-sphere. Instead of additional adaptive methods to control the learning rate in deep learning, the learning rate of TSGO is naturally determined by the angle $\theta $ as $\eta = \tan \theta $. Our numerical results reveal better convergence of TSGO in comparison to the off-the-shelf Adam.

Deep learning has achieved impressive prediction accuracies in a variety of scientific and industrial domains. However, the nested non-linear feature of deep learning makes the learning highly non-transparent, i.e., it is still unknown how the learning coordinates a huge number of parameters to achieve a decision making. To explain this hierarchical credit assignment, we propose a mean-field learning model by assuming that an ensemble of sub-networks, rather than a single network, are trained for a classification task. Surprisingly, our model reveals that apart from some deterministic synaptic weights connecting two neurons at neighboring layers, there exist a large number of connections that can be absent, and other connections can allow for a broad distribution of their weight values. Therefore, synaptic connections can be classified into three categories: very important ones, unimportant ones, and those of variability that may partially encode nuisance factors. Therefore, our model learns the credit assignment leading to the decision, and predicts an ensemble of sub-networks that can accomplish the same task, thereby providing insights toward understanding the macroscopic behavior of deep learning through the lens of distinct roles of synaptic weights.

Advances in deep neural networks (DNN) greatly bolster real-time detection of anomalous IoT data. However, IoT devices can barely afford complex DNN models due to limited computational power and energy supply. While one can offload anomaly detection tasks to the cloud, it incurs long delay and requires large bandwidth when thousands of IoT devices stream data to the cloud concurrently. In this paper, we propose an adaptive anomaly detection approach for hierarchical edge computing (HEC) systems to solve this problem. Specifically, we first construct three anomaly detection DNN models of increasing complexity, and associate them with the three layers of HEC from bottom to top, i.e., IoT devices, edge servers, and cloud. Then, we design an adaptive scheme to select one of the models based on the contextual information extracted from input data, to perform anomaly detection. The selection is formulated as a contextual bandit problem and is characterized by a single-step Markov decision process, with an objective of achieving high detection accuracy and low detection delay simultaneously. We evaluate our proposed approach using a real IoT dataset, and demonstrate that it reduces detection delay by 84% while maintaining almost the same accuracy as compared to offloading detection tasks to the cloud. In addition, our evaluation also shows that it outperforms other baseline schemes.

Autonomous underwater vehicle (AUV) plays an increasingly important role in ocean exploration. Existing AUVs are usually not fully autonomous and generally limited to pre-planning or pre-programming tasks. Reinforcement learning (RL) and deep reinforcement learning have been introduced into the AUV design and research to improve its autonomy. However, these methods are still difficult to apply directly to the actual AUV system because of the sparse rewards and low learning efficiency. In this paper, we proposed a deep interactive reinforcement learning method for path following of AUV by combining the advantages of deep reinforcement learning and interactive RL. In addition, since the human trainer cannot provide human rewards for AUV when it is running in the ocean and AUV needs to adapt to a changing environment, we further propose a deep reinforcement learning method that learns from both human rewards and environmental rewards at the same time. We test our methods in two path following tasks---straight line and sinusoids curve following of AUV by simulating in the Gazebo platform. Our experimental results show that with our proposed deep interactive RL method, AUV can converge faster than a DQN learner from only environmental reward. Moreover, AUV learning with our deep RL from both human and environmental rewards can also achieve a similar or even better performance than that with the deep interactive RL method and can adapt to the actual environment by further learning from environmental rewards.

The article below is a guest post by Nuance, a company focused on conversational AI. In this post, Nuance engineers describe their use of NVIDIA's automatic mixed precision to speed up their AI models in the healthcare industry. Nuance's ambient clinical intelligence (ACI) technology is an example of how it is accelerating development of solutions for urgent problems in the U.S. healthcare system by training its automatic speech recognition (ASR) and natural language processing (NLP) models using NVIDIA's Automatic Mixed Precision capabilities on Volta and Turing GPUs with Tensor Cores. ACI addresses what the World Medical Association calls a "pandemic of physician burnout" caused by huge amounts of electronic paperwork. Doctors spend two hours completing documentation for every hour they deliver care.