arXiv subjects

Zenan Ling

Publications and source records attributed to Zenan Ling.

35 records · Page 2Linked to original sources

arXiv2023

On the Equivalence between Implicit and Explicit Neural Networks: A High-dimensional Viewpoint

Implicit neural networks have demonstrated remarkable success in various tasks. However, there is a lack of theoretical analysis of the connections and differences between implicit and explicit networks. In this paper, we study high-dimensional implicit neural networks and provide the high dimensional equivalents for the corresponding conjugate kernels and neural tangent kernels. Built upon this, we establish the equivalence between implicit and explicit networks in high dimensions.

cs.LG↗

arXiv2023

Global Convergence of Over-parameterized Deep Equilibrium Models

A deep equilibrium model (DEQ) is implicitly defined through an equilibrium point of an infinite-depth weight-tied model with an input-injection. Instead of infinite computations, it solves an equilibrium point directly with root-finding and computes gradients with implicit differentiation. The training dynamics of over-parameterized DEQs are investigated in this study. By supposing a condition on the initial equilibrium point, we show that the unique equilibrium point always exists during the training process, and the gradient descent is proved to converge to a globally optimal solution at a linear convergence rate for the quadratic loss function. In order to show that the required initial condition is satisfied via mild over-parameterization, we perform a fine-grained analysis on random DEQs. We propose a novel probabilistic framework to overcome the technical difficulty in the non-asymptotic analysis of infinite-depth weight-tied models.

cs.LG↗

arXiv2022

A Roadmap for Big Model

With the rapid development of deep learning, training Big Models (BMs) for multiple downstream tasks becomes a popular paradigm. Researchers have achieved various outcomes in the construction of BMs and the BM application in many fields. At present, there is a lack of research work that sorts out the overall progress of BMs and guides the follow-up research. In this paper, we cover not only the BM technologies themselves but also the prerequisites for BM training and applications with BMs, dividing the BM review into four parts: Resource, Models, Key Technologies and Application. We introduce 16 specific BM-related topics in those four parts, they are Data, Knowledge, Computing System, Parallel Training System, Language Model, Vision Model, Multi-modal Model, Theory&Interpretability, Commonsense Reasoning, Reliability&Security, Governance, Evaluation, Machine Translation, Text Generation, Dialogue and Protein Research. In each topic, we summarize clearly the current studies and propose some future research directions. At the end of this paper, we conclude the further development of BMs in a more general view.

cs.LG↗

arXiv2021

Optimization Induced Equilibrium Networks

Implicit equilibrium models, i.e., deep neural networks (DNNs) defined by implicit equations, have been becoming more and more attractive recently. In this paper, we investigate an emerging question: can an implicit equilibrium model's equilibrium point be regarded as the solution of an optimization problem? To this end, we first decompose DNNs into a new class of unit layer that is the proximal operator of an implicit convex function while keeping its output unchanged. Then, the equilibrium model of the unit layer can be derived, named Optimization Induced Equilibrium Networks (OptEq), which can be easily extended to deep layers. The equilibrium point of OptEq can be theoretically connected to the solution of its corresponding convex optimization problem with explicit objectives. Based on this, we can flexibly introduce prior properties to the equilibrium points: 1) modifying the underlying convex problems explicitly so as to change the architectures of OptEq; and 2) merging the information into the fixed point iteration, which guarantees to choose the desired equilibrium point when the fixed point set is non-singleton. We show that deep OptEq outperforms previous implicit models even with fewer parameters. This work establishes the first step towards the optimization-guided design of deep models.

cs.LG↗

arXiv2020

Early Anomaly Detection and Location in Distribution Network: A Data-Driven Approach

The measurement data collected from the supervisory control and data acquisition (SCADA) system installed in distribution network can reflect the operational state of the network effectively. In this paper, a random matrix theory (RMT) based approach is developed for early anomaly detection and localization by using the data. For every feeder in the distribution network, a corresponding data matrix is formed. Based on the Marchenko-Pastur Law for the empirical spectral analysis of covariance `signal+noise' matrix, the linear eigenvalue statistics are introduced to indicate the anomaly, and the outliers and their corresponding eigenvectors are analyzed for locating the anomaly. As for the low observability feeders in the distribution network, an increasing data dimension algorithm is designed for the formulated low-dimensional matrices being more accurately analyzed. The developed approach can detect and localize the anomaly at an early stage, and it is robust to random disturbance and measurement error. Cases on Matpower simulation data and real SCADA data corroborate the feasibility of the approach.

stat.AP↗

arXiv2019

Spatio-Temporal Correlation Analysis of Online Monitoring Data for Anomaly Detection and Location in Distribution Networks

The online monitoring data in distribution networks contain rich information on the running states of the networks. By leveraging the data, this paper proposes a spatio-temporal correlation analysis approach for anomaly detection and location in distribution networks. First, spatio-temporal matrix for each feeder line in a distribution network is formulated and the spectrum of its covariance matrix is analyzed. The spectrum is complex and exhibits two aspects: 1) bulk, which arises from random noise or fluctuations and 2) spikes, which represents factors caused by anomaly signals or fault disturbances. Then, by connecting the estimation of the number of factors to the limiting empirical spectral density of covariance matrices of residuals, the spatio-temporal parameters are accurately estimated, during which free random variable techniques are used. Based on the estimators, anomaly indicators are designed to detect and locate the anomalies by exploring the variations of spatio-temporal correlations in the data. The proposed approach is sensitive to the anomalies and robust to random fluctuations, which makes it possible for detecting early anomalies and reducing false alarming rate. Case studies on both synthetic data and real-world online monitoring data verify the effectiveness and advantages of the proposed approach.

eess.SP↗

arXiv2019

Spectrum concentration in deep residual learning: a free probability approach

We revisit the initialization of deep residual networks (ResNets) by introducing a novel analytical tool in free probability to the community of deep learning. This tool deals with non-Hermitian random matrices, rather than their conventional Hermitian counterparts in the literature. As a consequence, this new tool enables us to evaluate the singular value spectrum of the input-output Jacobian of a fully-connected deep ResNet for both linear and nonlinear cases. With the powerful tool of free probability, we conduct an asymptotic analysis of the spectrum on the single-layer case, and then extend this analysis to the multi-layer case of an arbitrary number of layers. In particular, we propose to rescale the classical random initialization by the number of residual units, so that the spectrum has the order of $O(1)$, when compared with the large width and depth of the network. We empirically demonstrate that the proposed initialization scheme learns at a speed of orders of magnitudes faster than the classical ones, and thus attests a strong practical relevance of this investigation.

cs.LG↗

arXiv2019

Explaining AlphaGo: Interpreting Contextual Effects in Neural Networks

In this paper, we propose to disentangle and interpret contextual effects that are encoded in a pre-trained deep neural network. We use our method to explain the gaming strategy of the alphaGo Zero model. Unlike previous studies that visualized image appearances corresponding to the network output or a neural activation only from a global perspective, our research aims to clarify how a certain input unit (dimension) collaborates with other units (dimensions) to constitute inference patterns of the neural network and thus contribute to the network output. The analysis of local contextual effects w.r.t. certain input units is of special values in real applications. Explaining the logic of the alphaGo Zero model is a typical application. In experiments, our method successfully disentangled the rationale of each move during the Go game.

cs.CV↗

arXiv2018

A New Approach of Exploiting Self-Adjoint Matrix Polynomials of Large Random Matrices for Anomaly Detection and Fault Location

Synchronized measurements of a large power grid enable an unprecedented opportunity to study the spatialtemporal correlations. Statistical analytics for those massive datasets start with high-dimensional data matrices. Uncertainty is ubiquitous in a future's power grid. These data matrices are recognized as random matrices. This new point of view is fundamental in our theoretical analysis since true covariance matrices cannot be estimated accurately in a high-dimensional regime. As an alternative, we consider large-dimensional sample covariance matrices in the asymptotic regime to replace the true covariance matrices. The self-adjoint polynomials of large-dimensional random matrices are studied as statistics for big data analytics. The calculation of the asymptotic spectrum distribution (ASD) for such a matrix polynomial is understandably challenging. This task is made possible by a recent breakthrough in free probability, an active research branch in random matrix theory. This is the very reason why the work of this paper is inspired initially. The new approach is interesting in many aspects. The mathematical reason may be most critical. The real-world problems can be solved using this approach, however.

stat.AP↗

arXiv2018

Individual Recognition in Schizophrenia using Deep Learning Methods with Random Forest and Voting Classifiers: Insights from Resting State EEG Streams

Recently, there has been a growing interest in monitoring brain activity for individual recognition system. So far these works are mainly focussing on single channel data or fragment data collected by some advanced brain monitoring modalities. In this study we propose new individual recognition schemes based on spatio-temporal resting state Electroencephalography (EEG) data. Besides, instead of using features derived from artificially-designed procedures, modified deep learning architectures which aim to automatically extract an individual's unique features are developed to conduct classification. Our designed deep learning frameworks are proved of a small but consistent advantage of replacing the $softmax$ layer with Random Forest. Additionally, a voting layer is added at the top of designed neural networks in order to tackle the classification problem arisen from EEG streams. Lastly, various experiments are implemented to evaluate the performance of the designed deep learning architectures; Results indicate that the proposed EEG-based individual recognition scheme yields a high degree of classification accuracy: $81.6\%$ for characteristics in high risk (CHR) individuals, $96.7\%$ for clinically stable first episode patients with schizophrenia (FES) and $99.2\%$ for healthy controls (HC).

cs.CV↗

arXiv2018

A Novel Data-Driven Situation Awareness Approach for Future Grids--Using Large Random Matrices for Big Data Modeling

Data-driven approaches, when tasked with situation awareness, are suitable for complex grids with massive datasets. It is a challenge, however, to efficiently turn these massive datasets into useful big data analytics. To address such a challenge, this paper, based on random matrix theory (RMT), proposes a datadriven approach. The approach models massive datasets as large random matrices; it is model-free and requiring no knowledge about physical model parameters. In particular, the large data dimension N and the large time span T, from the spatial aspect and the temporal aspect respectively, lead to favorable results. The beautiful thing lies in that these linear eigenvalue statistics (LESs) built from data matrices follow Gaussian distributions for very general conditions, due to the latest breakthroughs in probability on the central limit theorems of those LESs. Numerous case studies, with both simulated data and field data, are given to validate the proposed new algorithms.

stat.ME↗

arXiv2018

An Accurate and Real-time Self-blast Glass Insulator Location Method Based On Faster R-CNN and U-net with Aerial Images

The location of broken insulators in aerial images is a challenging task. This paper, focusing on the self-blast glass insulator, proposes a deep learning solution. We address the broken insulators location problem as a low signal-noise-ratio image location framework with two modules: 1) object detection based on Fast R-CNN, and 2) classification of pixels based on U-net. A diverse aerial image set of some grid in China is tested to validated the proposed approach. Furthermore, a comparison is made among different methods and the result shows that our approach is accurate and real-time.

cs.CV↗

arXiv2017

A Data-driven Approach to Multi-event Analytics in Large-scale Power Systems Using Factor Model

Multi-event detection and recognition in real time is of challenge for a modern grid as its feature is usually non-identifiable. Based on factor model, this paper porposes a data-driven method as an alternative solution under the framework of random matrix theory. This method maps the raw data into a high-dimensional space with two parts: 1) the principal components (factors, mapping event signals); and 2) time series residuals (bulk, mapping white/non-Gaussian noises). The spatial information is extracted form factors, and the termporal infromation from residuals. Taking both spatial-tempral correlation into account, this method is able to reveal the multi-event: its components and their respective details, e.g., occurring time. Case studies based on the standard IEEE 118-bus system validate the proposed method.

stat.AP↗

arXiv2017

Spatio-Temporal Big Data Analysis for Smart Grids Based on Random Matrix Theory: A Comprehensive Study

A cornerstone of the smart grid is the advanced monitorability on its assets and operations. Increasingly pervasive installation of the phasor measurement units (PMUs) allows the so-called synchrophasor measurements to be taken roughly 100 times faster than the legacy supervisory control and data acquisition (SCADA) measurements, time-stamped using the global positioning system (GPS) signals to capture the grid dynamics. On the other hand, the availability of low-latency two-way communication networks will pave the way to high-precision real-time grid state estimation and detection, remedial actions upon network instability, and accurate risk analysis and post-event assessment for failure prevention. In this chapter, we firstly modelling spatio-temporal PMU data in large scale grids as random matrix sequences. Secondly, some basic principles of random matrix theory (RMT), such as asymptotic spectrum laws, transforms, convergence rate and free probability, are introduced briefly in order to the better understanding and application of RMT technologies. Lastly, the case studies based on synthetic data and real data are developed to evaluate the performance of the RMT-based schemes in different application scenarios (i.e., state evaluation and situation awareness).

stat.AP↗

arXiv2017

Massive Streaming PMU Data Modeling and Analytics in Smart Grid State Evaluation Based on Multiple High-Dimensional Covariance Tests

The analogous deployment of phase measurement units (PMUs), the increase of data quantum and the deregulation of energy market, all call for the robust state evaluation in large scale power systems. Implementing model based estimators is impractical because of the complexity scale of solving the high dimension power flow equations. In this paper, we first represent massive streaming PMU data as big random matrix flow. By exploiting the variations in the covariance matrix of the massive streaming PMU data, a novel power state evaluation algorithm is then developed based on the multiple high dimensional covariance matrix tests. The proposed test statistic is flexible and nonparametric, which assumes no specific parameter distribution or dimension structure for the PMU data. Besides, it can jointly reveal the relative magnitude, duration and location of a system event. For the sake of practical application, we reduce the computation of the proposed test statistic from $O(\varepsilon n_g^4)$ to $O(ηn_g^2)$ by principal component calculation and redundant computation elimination. The novel algorithm is numerically evaluated utilizing the IEEE 30-, 118-bus system, a Polish 2383-bus system, and a real 34-PMU system. The case studies illustrate and verify the superiority of proposed state evaluation indicator.

stat.AP↗

arXiv2016

A Novel Approach for Big Data Analytics in Future Grids Based on Free Probability

Based on the random matrix model, we can build statistical models using massive datasets across the power grid, and employ hypothesis testing for anomaly detection. First, the aim of this paper is to make the first attempt to apply the recent free probability result in extracting big data analytics, in particular data fusion. The nature of this work is basic in that new algorithms and analytics tools are proposed to pave the way for the future's research. Second, using the new analytic tool, we are able to make some discovery related to anomaly detection that is very difficult for other approaches. To our best knowledge, there is no similar report in the literature. Third, both linear and nonlinear polynomials of large random matrices can be handled in this new framework. Simulations demonstrate the following: Compared with the linearity, nonlinearity is more flexible in problem modeling and closer to the nature of the reality. In some sense, some other nonlinear matrix polynomials may be more effective for the power grid

stat.AP↗

arXiv2016

Designing for Situation Awareness of Future Power Grids: An Indicator System Based on Linear Eigenvalue Statistics of Large Random Matrices

Future power grids are fundamentally different from current ones, both in size and in complexity; this trend imposes challenges for situation awareness (SA) based on classical indicators, which are usually model-based and deterministic. As an alternative, this paper proposes a statistical indicator system based on linear eigenvalue statistics (LESs) of large random matrices: 1) from a data modeling viewpoint, we build, starting from power flows equations, the random matrix models (RMMs) only using the real-time data flow in a statistical manner; 2) for a data analysis that is fully driven from RMMs, we put forward the high-dimensional indicators, called LESs that have some unique statistical features such as Gaussian properties; and 3) we develop a three-dimensional (3D) power-map to visualize the system, respectively, from a high-dimensional viewpoint and a low-dimensional one. Therefore, a statistical methodology of SA is employed; it conducts SA with a model-free and data-driven procedure, requiring no knowledge of system topologies, units operation/control models, causal relationship, etc. This methodology has numerous advantages, such as sensitivity, universality, speed, and flexibility. In particular, its robustness against bad data is highlighted, with potential advantages in cyber security. The theory of big data based stability for on-line operations may prove feasible along with this line of work, although this critical development will be reported elsewhere.

stat.ME↗