arXiv subjects

Shen Wang

Publications and source records attributed to Shen Wang.

At least 127 records · Page 7Linked to original sources

arXiv2020

Attentional Graph Convolutional Networks for Knowledge Concept Recommendation in MOOCs in a Heterogeneous View

Massive open online courses are becoming a modish way for education, which provides a large-scale and open-access learning opportunity for students to grasp the knowledge. To attract students' interest, the recommendation system is applied by MOOCs providers to recommend courses to students. However, as a course usually consists of a number of video lectures, with each one covering some specific knowledge concepts, directly recommending courses overlook students'interest to some specific knowledge concepts. To fill this gap, in this paper, we study the problem of knowledge concept recommendation. We propose an end-to-end graph neural network-based approach calledAttentionalHeterogeneous Graph Convolutional Deep Knowledge Recommender(ACKRec) for knowledge concept recommendation in MOOCs. Like other recommendation problems, it suffers from sparsity issues. To address this issue, we leverage both content information and context information to learn the representation of entities via graph convolution network. In addition to students and knowledge concepts, we consider other types of entities (e.g., courses, videos, teachers) and construct a heterogeneous information network to capture the corresponding fruitful semantic relationships among different types of entities and incorporate them into the representation learning process. Specifically, we use meta-path on the HIN to guide the propagation of students' preferences. With the help of these meta-paths, the students' preference distribution with respect to a candidate knowledge concept can be captured. Furthermore, we propose an attention mechanism to adaptively fuse the context information from different meta-paths, in order to capture the different interests of different students. The promising experiment results show that the proposedACKRecis able to effectively recommend knowledge concepts to students pursuing online learning in MOOCs.

cs.LG↗

arXiv2020

Generative Temporal Link Prediction via Self-tokenized Sequence Modeling

We formalize networks with evolving structures as temporal networks and propose a generative link prediction model, Generative Link Sequence Modeling (GLSM), to predict future links for temporal networks. GLSM captures the temporal link formation patterns from the observed links with a sequence modeling framework and has the ability to generate the emerging links by inferring from the probability distribution on the potential future links. To avoid overfitting caused by treating each link as a unique token, we propose a self-tokenization mechanism to transform each raw link in the network to an abstract aggregation token automatically. The self-tokenization is seamlessly integrated into the sequence modeling framework, which allows the proposed GLSM model to have the generalization capability to discover link formation patterns beyond raw link sequences. We compare GLSM with the existing state-of-art methods on five real-world datasets. The experimental results demonstrate that GLSM obtains future positive links effectively in a generative fashion while achieving the best performance (2-10\% improvements on AUC) among other alternatives.

cs.LG↗

arXiv2020

State Estimation in Water Distribution Networks through a New Successive Linear Approximation

State estimation (SE) of water distribution networks (WDNs) is difficult to solve due to nonlinearity/nonconvexity of water flow models, uncertainties from parameters and demands, lack of redundancy of measurements, and inaccurate flow and pressure measurements. This paper proposes a new, scalable successive linear approximation to solve the SE problem in WDNs. The approach amounts to solving either a sequence of linear or quadratic programs---depending on the operators' objectives. The proposed successive linear approximation offers a seamless way of dealing with valve/pump model nonconvexities, is different than a first order Taylor series linearization, and can incorporate with robust uncertainty modeling. Two simple testcases are adopted to illustrate the effectiveness of proposed approach using head measurements at select nodes.

eess.SY↗

arXiv2020

Chip Placement with Deep Reinforcement Learning

In this work, we present a learning-based approach to chip placement, one of the most complex and time-consuming stages of the chip design process. Unlike prior methods, our approach has the ability to learn from past experience and improve over time. In particular, as we train over a greater number of chip blocks, our method becomes better at rapidly generating optimized placements for previously unseen chip blocks. To achieve these results, we pose placement as a Reinforcement Learning (RL) problem and train an agent to place the nodes of a chip netlist onto a chip canvas. To enable our RL policy to generalize to unseen blocks, we ground representation learning in the supervised task of predicting placement quality. By designing a neural architecture that can accurately predict reward across a wide variety of netlists and their placements, we are able to generate rich feature embeddings of the input netlists. We then use this architecture as the encoder of our policy and value networks to enable transfer learning. Our objective is to minimize PPA (power, performance, and area), and we show that, in under 6 hours, our method can generate placements that are superhuman or comparable on modern accelerator netlists, whereas existing baselines require human experts in the loop and take several weeks.

cs.LG↗

arXiv2020

BusTime: Which is the Right Prediction Model for My Bus Arrival Time?

With the rise of big data technologies, many smart transportation applications have been rapidly developed in recent years including bus arrival time predictions. This type of applications help passengers to plan trips more efficiently without wasting unpredictable amount of waiting time at bus stops. Many studies focus on improving the prediction accuracy of various machine learning and statistical models, while much less work demonstrate their applicability of being deployed and used in realistic urban settings. This paper tries to fill this gap by proposing a general and practical evaluation framework for analysing various widely used prediction models (i.e. delay, k-nearest-neighbour, kernel regression, additive model, and recurrent neural network using long short term memory) for bus arrival time. In particular, this framework contains a raw bus GPS data pre-processing method that needs much less number of input data points while still maintain satisfactory prediction results. This pre-processing method enables various models to predict arrival time at bus stops only, by using a KD-tree based nearest point search method. Based on this framework, using raw bus GPS dataset in different scales from the city of Dublin, Ireland, we also present preliminary results for city managers by analysing the practical strengths and weaknesses in both training and predicting stages of commonly used prediction models.

eess.SP↗

arXiv2020

Computing Lipschitz Constants for Hydraulic Models of Water Distribution Networks

Drinking water distribution networks (WDN) are large-scale, dynamic systems spanning large geographic areas. Water networks include various components such as junctions, reservoirs, tanks, pipes, pumps, and valves. Hydraulic models for these components depicting mass and energy balance form nonlinear algebraic differential equations (NDAE). While control theoretic studies have been thoroughly explored for other complex infrastructure such as power and transportation systems, little is understood or even investigated for feedback control and state estimation problems for the NDAE models of WDN. The objective of this letter is to showcase a complete NDAE model of WDN followed by computing Lipschitz constants of the vector-valued nonlinearity in that model. The computation of Lipschitz constants of hydraulic models is crucial as it paves the way to apply a plethora of control-theoretic studies for water system applications. In particular, the computation of Lipschitz constant is explored through closed-form, analytical expressions as well as via numerical methods. Case studies reveal how such computations fare against each other for various water networks.

math.OC↗

arXiv2020

Segmentation-based Method combined with Dynamic Programming for Brain Midline Delineation

The midline related pathological image features are crucial for evaluating the severity of brain compression caused by stroke or traumatic brain injury (TBI). The automated midline delineation not only improves the assessment and clinical decision making for patients with stroke symptoms or head trauma but also reduces the time of diagnosis. Nevertheless, most of the previous methods model the midline by localizing the anatomical points, which are hard to detect or even missing in severe cases. In this paper, we formulate the brain midline delineation as a segmentation task and propose a three-stage framework. The proposed framework firstly aligns an input CT image into the standard space. Then, the aligned image is processed by a midline detection network (MD-Net) integrated with the CoordConv Layer and Cascade AtrousCconv Module to obtain the probability map. Finally, we formulate the optimal midline selection as a pathfinding problem to solve the problem of the discontinuity of midline delineation. Experimental results show that our proposed framework can achieve superior performance on one in-house dataset and one public dataset.

eess.IV↗

arXiv2020

New Insights on One-Sided Lipschitz and Quadratically Inner-Bounded Nonlinear Dynamic Systems

Nonlinear dynamic systems can be classified into various classes depending on the modeled nonlinearity. These classes include Lipschitz, bounded Jacobian, one-sided Lipschitz (OSL), and quadratically inner-bounded (QIB). Such classes essentially yield bounding constants characterizing the nonlinearity. This is then used to design observers and controllers through Riccati equations or matrix inequalities. While analytical expressions for bounding constants of Lipschitz and bounded Jacobian nonlinearity are studied in the literature, OSL and QIB classes are not thoroughly analyzed---computationally or analytically. In short, this paper develops analytical expressions of OSL and QIB bounding constants. These expressions are posed as constrained maximization problems, which can be solved via various optimization algorithms. This paper also presents a novel insight particularly on QIB function set: any function that is QIB turns out to be also Lipschitz continuous.

math.OC↗

arXiv2020

A New Derivative-Free Linear Approximation for Solving the Network Water Flow Problem with Convergence Guarantees

Addressing challenges in urban water infrastructure systems including aging infrastructure, supply uncertainty, extreme events, and security threats, depend highly on water distribution networks modeling emphasizing the importance of realistic assumptions, modeling complexities, and scalable solutions. In this study, we propose a derivative-free, linear approximation for solving the network water flow problem (WFP). The proposed approach takes advantage of the special form of the nonlinear head loss equations and, after the transformation of variables and constraints, the WFP reduces to a linear optimization problem that can be efficiently solved by modern linear solvers. Ultimately, the proposed approach amounts to solving a series of linear optimization problems. We demonstrate the proposed approach through several case studies and show that the approach can model arbitrary network topologies and various types of valves and pumps, thus providing modeling flexibility. Under mild conditions, we show that the proposed linear approximation converges. We provide sensitivity analysis and discuss in detail the current limitations of our approach and suggest solutions to overcome these. All the codes, tested networks, and results are freely available on Github for research reproducibility.

math.OC↗

arXiv2019

A PRESTO-based Parallel Pulsar Search Pipeline Used for FAST Drift Scan Data

We developed a pulsar search pipeline based on PRESTO (PulsaR Exploration and Search Toolkit). This pipeline simply runs dedispersion, FFT (Fast Fourier Transformation), and acceleration search in process-level parallel to shorten the processing time. With two parallel strategies, the pipeline can highly shorten the processing time in both the normal searches or acceleration searches. This pipeline was first tested with PMPS (Parkes Multibeam Pulsar Survery) data and discovered two new faint pulsars. Then, it was successfully used in processing the FAST (Five-hundred-meter Aperture Spherical radio Telescope) drift scan data with tens of new pulsar discoveries up to now. The pipeline is only CPU-based and can be easily and quickly deployed in computing nodes for testing purposes or data processes.

astro-ph.IM↗

arXiv2019

Receding Horizon Control for Drinking Water Networks: The Case for Geometric Programming

Optimal, network-driven control of Water Distribution Networks (WDN) is very difficult: valve and pump models form non-trivial, combinatorial logic; hydraulic models are nonconvex; water demand patterns are uncertain; and WDN are naturally large-scale. Prior research on control of WDN addressed major research challenges, yet either (i) adopted simplified hydraulic models, WDN topologies, and rudimentary valve/pump modeling or (ii) used mixed-integer, nonconvex optimization to solve WDN control problems. The objective of this paper is to develop tractable computational algorithms to manage WDN operation, while considering arbitrary topology, flow direction, an abundance of valve types, control objectives, hydraulic models, and operational constraints---all while only using convex, continuous optimization. Specifically, we propose new Geometric Programming (GP)-based Model Predictive Control (MPC) algorithms, designed to solve the water flow equations and obtain WDN controls, i.e., pump/valve schedules alongside heads and flows. The proposed approach amounts to solving a series of convex optimization problems that graciously scale to large networks. The proposed approach is tested using a 126-node network with many valves and pumps and shown to outperform traditional, rule-based control. The developed GP-based MPC algorithms, as well as the numerical test results are all included on Github.

eess.SY↗

arXiv2019

Probing the emission states of PSR J1107-5907

The emission from PSR J1107-5907 is erratic. Sometimes the radio pulse is undetectable, at other times the pulsed emission is weak, and for short durations the emission can be very bright. In order to improve our understanding of these state changes, we have identified archival data sets from the Parkes radio telescope in which the bright emission is present, and find that the emission never switches from the bright state to the weak state, but instead always transitions to the off state. Previous work had suggested the identification of the off state as an extreme manifestation of the weak state. However, the connection between the off and bright emission reported here suggests that the emission can be interpreted as undergoing only two emission states: a bursting state consisting of both bright pulses and nulls as well as the weak-emission state.

astro-ph.HE↗

arXiv2019

Heterogeneous Graph Matching Networks

Information systems have widely been the target of malware attacks. Traditional signature-based malicious program detection algorithms can only detect known malware and are prone to evasion techniques such as binary obfuscation, while behavior-based approaches highly rely on the malware training samples and incur prohibitively high training cost. To address the limitations of existing techniques, we propose MatchGNet, a heterogeneous Graph Matching Network model to learn the graph representation and similarity metric simultaneously based on the invariant graph modeling of the program's execution behaviors. We conduct a systematic evaluation of our model and show that it is accurate in detecting malicious program behavior and can help detect malware attacks with less false positives. MatchGNet outperforms the state-of-the-art algorithms in malware detection by generating 50% less false positives while keeping zero false negatives.

cs.CR↗

arXiv2019

Energy Crowdsourcing and Peer-to-Peer Energy Trading in Blockchain-Enabled Smart Grids

The power grid is rapidly transforming, and while recent grid innovations increased the utilization of advanced control methods, the next-generation grid demands technologies that enable the integration of distributed energy resources (DERs)---and consumers that both seamlessly buy and sell electricity. This paper develops an optimization model and blockchain-based architecture to manage the operation of crowdsourced energy systems (CES), with peer-to-peer (P2P) energy trading transactions. An operational model of CESs in distribution networks is presented considering various types of energy trading transactions and crowdsourcees. Then, a two-phase operation algorithm is presented: Phase I focuses on the day-ahead scheduling of generation and controllable DERs, whereas Phase II is developed for hour-ahead or real-time operation of distribution networks. The developed approach supports seamless P2P energy trading between individual prosumers and/or the utility. The presented operational model can also be used to operate islanded microgrids. The CES framework and the operation algorithm are then prototyped through an efficient blockchain implementation, namely the IBM Hyperledger Fabric. This implementation allows the system operator to manage the network users to seamlessly trade energy. Case studies and prototype illustration are provided.

eess.SY↗

arXiv2019

Adversarial Defense Framework for Graph Neural Network

Graph neural network (GNN), as a powerful representation learning model on graph data, attracts much attention across various disciplines. However, recent studies show that GNN is vulnerable to adversarial attacks. How to make GNN more robust? What are the key vulnerabilities in GNN? How to address the vulnerabilities and defense GNN against the adversarial attacks? In this paper, we propose DefNet, an effective adversarial defense framework for GNNs. In particular, we first investigate the latent vulnerabilities in every layer of GNNs and propose corresponding strategies including dual-stage aggregation and bottleneck perceptron. Then, to cope with the scarcity of training data, we propose an adversarial contrastive learning method to train the GNN in a conditional GAN manner by leveraging the high-level graph representation. Extensive experiments on three public datasets demonstrate the effectiveness of DefNet in improving the robustness of popular GNN variants, such as Graph Convolutional Network and GraphSAGE, under various types of adversarial attacks.

cs.LG↗

arXiv2019

Attentional Heterogeneous Graph Neural Network: Application to Program Reidentification

Program or process is an integral part of almost every IT/OT system. Can we trust the identity/ID (e.g., executable name) of the program? To avoid detection, malware may disguise itself using the ID of a legitimate program, and a system tool (e.g., PowerShell) used by the attackers may have the fake ID of another common software, which is less sensitive. However, existing intrusion detection techniques often overlook this critical program reidentification problem (i.e., checking the program's identity). In this paper, we propose an attentional heterogeneous graph neural network model (DeepHGNN) to verify the program's identity based on its system behaviors. The key idea is to leverage the representation learning of the heterogeneous program behavior graph to guide the reidentification process. We formulate the program reidentification as a graph classification problem and develop an effective attentional heterogeneous graph embedding algorithm to solve it. Extensive experiments --- using real-world enterprise monitoring data and real attacks --- demonstrate the effectiveness of DeepHGNN across multiple popular metrics and the robustness to the normal dynamic changes like program version upgrades.

cs.CR↗

arXiv2019

The First Pulsar Discovered by FAST

To assist with the commissioning (Jiang et al. 2019) of the Five-hundred-meter Aperture Spherical radio Telescope (FAST), we performed a pulsar search, with the primary goal of developing and testing the pulsar data acquisition and processing pipelines. We tested and used three pipelines, two (P1 and P2 hereafter) searched for the periodic signature of pulsars whereas the other one was used to search for bright single pulses (P3 hereafter). A pulsar candidate was discovered in the observation on the 22nd August, 2017, and later confirmed by the Parkes radio telescope on the 10th September, 2017. This pulsar, named PSR J1900-0134, was the first pulsar discovered by FAST. The pulsar has a pulse period of 1.8 s and a dispersion measure (DM) of 188\,pc\,cm$^{-3}$.

astro-ph.IM↗

arXiv2019

Flux density measurements for 32 pulsars in the 20 cm band

Flux density measurements provide fundamental observational parameters that describe a pulsar. In the current pulsar catalogue, 27% of radio pulsars have no flux density measurement in the 20 cm observing band. Here, we present the first measurements of the flux densities in this band for 32 pulsars observed using the Parkes radio telescope and provide updated pulse profiles for these pulsars. We have used both archival and new observations to make these measurements. Various schemes exist for measuring flux densities. We show how the flux densities measured vary between these methods and how the presence of radio-frequency-interference will bias flux density measurements

astro-ph.GA↗