A deep neural network is a hierarchical nonlinear model transforming input signals to output signals. Its input-output relation is considered to be stochastic, being described for a given input by a parameterized conditional probability distribution of outputs. The space of parameters consisting of weights and biases is a Riemannian manifold, where the metric is defined by the Fisher information matrix. The natural gradient method uses the steepest descent direction in a Riemannian manifold, so it is effective in learning, avoiding plateaus. It requires inversion of the Fisher information matrix, however, which is practically impossible when the matrix has a huge number of dimensions. Many methods for approximating the natural gradient have therefore been introduced. The present paper uses statistical neurodynamical method to reveal the properties of the Fisher information matrix in a net of random connections under the mean field approximation. We prove that the Fisher information matrix is unit-wise block diagonal supplemented by small order terms of off-block-diagonal elements, which provides a justification for the quasi-diagonal natural gradient method by Y. Ollivier. A unitwise block-diagonal Fisher metrix reduces to the tensor product of the Fisher information matrices of single units. We further prove that the Fisher information matrix of a single unit has a simple reduced form, a sum of a diagonal matrix and a rank 2 matrix of weight-bias correlations. We obtain the inverse of Fisher information explicitly. We then have an explicit form of the natural gradient, without relying on the numerical matrix inversion, which drastically speeds up stochastic gradient learning.

Abstract--We propose channel charting (CC), a novel framework in which a multi-antenna network element learns a chart of the radio geometry in its surrounding area. The channel chart captures the local spatial geometry of the area so that points that are close in space will also be close in the channel chart and vice versa. CC works in a fully unsupervised manner, i.e., learning is only based on channel state information (CSI) that is passively collected at a single point in space, but from multiple transmit locations in the area over time. The method then extracts channel features that characterize large-scale fading properties of the wireless channel. Finally, the channel charts are generated with tools from dimensionality reduction, manifold learning, and deep neural networks. The network element performing CC may be, for example, a multi-antenna base-station in a cellular system and the charted area in the served cell. Logical relationships related to the position and movement of a transmitter, e.g., a user equipment (UE), in the cell can then be directly deduced from comparing measured radio channel characteristics to the channel chart. The unsupervised nature of CC enables a range of new applications in UE localization, network planning, user scheduling, multipoint connectivity, handover, cell search, user grouping, and other cognitive tasks that rely on CSI and UE movement relative to the base-station, without the need of information from global navigation satellite systems. UTURE wireless communication systems must sustain a massive increase in traffic volumes, number of terminals, and reliability/latency requirements [2], [3]. C. Studer, S. Medjkouh, and E. Gönültaş are with the School of Electrical and Computer Engineering, Cornell University, Ithaca, NY; email: studer@cornell.edu, T. Goldstein is with the Department of Computer Science, University of Maryland, College Park, MD; email: tomg@cs.umd.edu O. Tirkkonen was a visiting professor at the School of Electrical and Computer Engineering, Cornell University, Ithaca, NY, and is now at the School of Electrical Engineering, Aalto University, Finland; email: olav.tirkkonen@aalto.fi The work of CS, SM, and EG was supported in part by Xilinx Inc., and by the US NSF under grants ECCS-1408006, CCF-1535897, CAREER CCF-1652065, and CNS-1717559.

The ability to witness non-local correlations lies at the core of foundational aspects of quantum mechanics and its application in the processing of information. Commonly, this is achieved via the violation of Bell inequalities. Unfortunately, however, their systematic derivation quickly becomes unfeasible as the scenario of interest grows in complexity. To cope with that, we propose here a machine learning approach for the detection and quantification of non-locality. It consists of an ensemble of multilayer perceptrons blended with genetic algorithms achieving a high performance in a number of relevant Bell scenarios. Our results offer a novel method and a proof-of-principle for the relevance of machine learning for understanding non-locality.

The training of deep neural network classifiers results in decision boundaries which geometry is still not well understood. This is in direct relation with classification problems such as so called adversarial examples. We introduce zoNNscan, an index that is intended to inform on the boundary uncertainty (in terms of the presence of other classes) around one given input datapoint. It is based on confidence entropy, and is implemented through sampling in the multidimensional ball surrounding that input. We detail the zoNNscan index, give an algorithm for approximating it, and finally illustrate its benefits on four applications, including two important problems for the adoption of deep networks in critical systems: adversarial examples and corner case inputs. We highlight that zoNNscan exhibits significantly higher values than for standard inputs in those two problem classes.

Population synthesis is concerned with the generation of synthetic yet realistic representations of populations. It is a fundamental problem in the modeling of transport where the synthetic populations of micro agents represent a key input to most agent-based models. In this paper, a new methodological framework for how to grow pools of micro agents is presented. This is accomplished by adopting a deep generative modeling approach from machine learning based on a Variational Autoencoder (VAE) framework. Compared to the previous population synthesis approaches based on Iterative Proportional Fitting (IPF), Markov Chain Monte Carlo (MCMC) sampling or traditional generative models, the proposed method allows unparalleled scalability with respect to the number and types of attributes. In contrast to the approaches that rely on approximating the joint distribution in the observed data space, VAE learns its compressed latent representation. The advantage of the compressed representation is that it avoids the problem of the generated samples being trapped in local minima when the number of attributes becomes large. The problem is illustrated using the Danish National Travel Survey data, where the Gibbs sampler fails to generate a population with 21 attributes (corresponding to the 121-dimensional joint distribution). At the same time, VAE shows acceptable performance when 47 attributes (corresponding to the 357-dimensional joint distribution) are used. Moreover, VAE allows for growing agents that are virtually different from those in the original data but have similar statistical properties and correlation structure. The presented approach will help modelers to generate better and richer populations with a high level of detail, including smaller zones, personal details and travel preferences.

Convolutional Neural Networks (CNNs) have demonstrated their superiority in image classification, and evolutionary computation (EC) methods have recently been surging to automatically design the architectures of CNNs to save the tedious work of manually designing CNNs. In this paper, a new hybrid differential evolution (DE) algorithm with a newly added crossover operator is proposed to evolve the architectures of CNNs of any lengths, which is named DECNN. There are three new ideas in the proposed DECNN method. Firstly, an existing effective encoding scheme is refined to cater for variable-length CNN architectures; Secondly, the new mutation and crossover operators are developed for variable-length DE to optimise the hyperparameters of CNNs; Finally, the new second crossover is introduced to evolve the depth of the CNN architectures. The proposed algorithm is tested on six widely-used benchmark datasets and the results are compared to 12 state-of-the-art methods, which shows the proposed method is vigorously competitive to the state-of-the-art algorithms. Furthermore, the proposed method is also compared with a method using particle swarm optimisation with a similar encoding strategy named IPPSO, and the proposed DECNN outperforms IPPSO in terms of the accuracy.

Common-sense and background knowledge is required to understand natural language, but in most neural natural language understanding (NLU) systems, this knowledge must be acquired from training corpora during learning, and then it is static at test time. We introduce a new architecture for the dynamic integration of explicit background knowledge in NLU models. A general-purpose reading module reads background knowledge in the form of free-text statements (together with task-specific text inputs) and yields refined word representations to a task-specific NLU architecture that reprocesses the task inputs with these representations. Experiments on document question answering (DQA) and recognizing textual entailment (RTE) demonstrate the effectiveness and flexibility of the approach. Analysis shows that our model learns to exploit knowledge in a semantically appropriate way.

With the development of information technology, there is an explosive growth in the number of online comment concerning news, blogs and so on. The massive comments are overloaded, and often contain some misleading and unwelcome information. Therefore, it is necessary to identify high-quality comments and filter out low-quality comments. In this work, we introduce a novel task: high-quality comment identification (HQCI), which aims to automatically assess the quality of online comments. First, we construct a news comment corpus, which consists of news, comments, and the corresponding quality label. Second, we analyze the dataset, and find the quality of comments can be measured in three aspects: informativeness, consistency, and novelty. Finally, we propose a novel multi-target text matching model, which can measure three aspects by referring to the news and surrounding comments. Experimental results show that our method can outperform various baselines by a large margin on the news dataset.

We propose a framework to automatically generate descriptive comments for source code blocks. While this problem has been studied by many researchers previously, their methods are mostly based on fixed template and achieves poor results. Our framework does not rely on any template, but makes use of a new recursive neural network called Code-RNN to extract features from the source code and embed them into one vector. When this vector representation is input to a new recurrent neural network (Code-GRU), the overall framework generates text descriptions of the code with accuracy (Rouge-2 value) significantly higher than other learning-based approaches such as sequence-to-sequence model. The Code-RNN model can also be used in other scenario where the representation of code is required.

In this paper, we study the product title summarization problem in E-commerce applications for display on mobile devices. Comparing with conventional sentence summarization, product title summarization has some extra and essential constraints. For example, factual detail errors or loss of the key information are intolerable for E-commerce applications. Therefore, we abstract two more constraints for product title summarization: (i) do not introduce irrelevant information; (ii) retain the key information (e.g., brand name and commodity name). To address these issues, we propose a novel multi-source pointer network by adding a new knowledge encoder for pointer network. The first constraint is handled by pointer mechanism, generating the short title by copying words from the source title. For the second constraint, we restore the key information by copying words from the knowledge encoder with the help of the soft gating mechanism. For evaluation, we build a large collection of real-world product titles along with human-written short titles. Experimental results demonstrate that our model significantly outperforms the other baselines.Finally, online deployment of our proposed model has yielded a significant business impact, as measured by the click-through rate.