A NUMA-Aware Parallel Truss Decomposition Algorithm for Large Scale Graphs

We present PKT, a new shared-memory parallel algorithm and OpenMP implementation for the truss decomposition of large sparse graphs. A k-truss is a dense subgraph definition that can be considered a relaxation of a clique. Truss decomposition refers to a partitioning of all the edges in the graph based on their k-truss membership. The truss decomposition of a graph has many applications. We show that our new approach PKT consistently outperforms other truss decomposition approaches for a collection of large sparse graphs and on a 24-core shared-memory server. PKT is based on a recently proposed algorithm for k-core decomposition.

Graph-structured analytics has been widely adopted in a number of big data applications such as social computation, web-search and recommendation systems. Though much prior research focuses on scaling graph-analytics on distributed environments, the strong desire on performance per core, dollar and joule has generated considerable interests of processing large-scale graphs on a single server-class machine, which may have several terabytes of RAM and 80 or more cores. However, prior graph-analytics systems are largely neutral to NUMA characteristics and thus have suboptimal performance. This paper presents a detailed study of NUMA characteristics and their impact on the efficiency of graph-analytics. Our study uncovers two insights: 1) either random or interleaved allocation of graph data will significantly hamper data locality and parallelism; 2) sequential inter-node (i.e., remote) memory accesses have much higher bandwidth than both intra- and inter-node random ones. Based on them, this paper describes Polymer, a NUMA-aware graph-analytics system on multicore with two key design decisions. First, Polymer differentially allocates and places topology data, application-defined data and mutable runtime states of a graph system according to their access patterns to minimize remote accesses. Second, for some remaining random accesses, Polymer carefully converts random remote accesses into sequential remote accesses, by using lightweight replication of vertices across NUMA nodes. To improve load balance and vertex convergence, Polymer is further built with a hierarchical barrier to boost parallelism and locality, an edge-oriented balanced partitioning for skewed graphs, and adaptive data structures according to the proportion of active vertices. A detailed evaluation on an 80-core machine shows that Polymer often outperforms the state-of-the-art single-machine graph-analytics systems, including Ligra, X-Stream and Galois, for a set of popular real-world and synthetic graphs.

Performance degradation due to nonuniform data access latencies has worsened on NUMA systems and can now be felt on-chip in manycore processors. Distributing data across NUMA nodes and manycore processor caches is necessary to reduce the impact of nonuniform latencies. However, techniques for distributing data are error-prone and fragile and require low-level architectural knowledge. Existing task scheduling policies favor quick load-balancing at the expense of locality and ignore NUMA node/manycore cache access latencies while scheduling. Locality-aware scheduling, in conjunction with or as a replacement for existing scheduling, is necessary to minimize NUMA effects and sustain performance. We present a data distribution and locality-aware scheduling technique for task-based OpenMP programs executing on NUMA systems and manycore processors. Our technique relieves the programmer from thinking of NUMA system/manycore processor architecture details by delegating data distribution to the runtime system and uses task data dependence information to guide the scheduling of OpenMP tasks to reduce data stall times. We demonstrate our technique on a four-socket AMD Opteron machine with eight NUMA nodes and on the TILEPro64 processor and identify that data distribution and locality-aware task scheduling improve performance up to 69% for scientific benchmarks compared to default policies and yet provide an architecture-oblivious approach for programmers.

The truss decomposition provides a popular model for discovering cohesive communities in a given network (graph). The problem has been well studied in sequential, shared memory and MapReduce settings. We study the problem on distributed memory systems. Our work builds on two prior algorithms. The first algorithm is optimized in terms of the computational load and communication volume, but it involves a large number of iterations, leading to high load imbalance and synchronization costs. The second algorithm significantly reduces the number of iterations, but at the cost of increasing the load and the volume. We design an algorithm that offers a tradeoff between the two extremes, with the number of iterations being close to that of the second algorithm and load/volume being close to that of the first. We develop an efficient distributed (MPI) implementation based on the new algorithm. We present an experimental evaluation on large real-world graphs. The evaluation shows that the new algorithm outperforms the two prior algorithms on large system sizes with the performance gain ranging up to 2x.

Parallel K-Tree: A multicore, multinode solution to extreme clustering

Non-intrusive load monitoring (NILM) is to estimate individual appliance’s power consumption from aggregated smart meter data, which is useful for optimized energy management and provisioning of customized services. While deep learning (DL) has achieved state-of-the-art NILM performance, it is still constrained by the dependency on large amounts of data and intensive computations on training. In this paper, we propose a pre-training approach to address the generalization of DL models for NILM. We develop a meta-learning based approach and an ensemble learning based approach, which pre-train a base model and then fine-tune it with few-short learning when applied to an unknown dataset. The models are validated with two real-world datasets and shown to achieve a superior transferability performance compared with traditional DL and transfer learning methods.

Injecting purpose and trust into data anonymisation

Visualization of remote data needs software that is fundamentally designed for distributed processing and high-speed networks. Motivated by and following the requirements of a distributed visualization application, this article describes the design and implementation of a fast and configurable remote data access system called eavivdata. Because wide-area networks may have a high latency, the remote data access system uses an architecture that effectively hides latency. Four remote data access architectures are analyzed and the results show that an architecture that combines bulk and pipeline processing is the best solution for high-throughput remote data access. The resulting system, also supporting high-speed transport protocols and configurable remote operations, is up to 400 times faster than a comparable existing remote data access system.

We present a novel architecture of a communication engine for non-coherent distributed shared memory systems. The shared memory is composed by a set of nodes exporting their memory. Remote memory access is possible by forwarding local load or store transactions to remote nodes. No software layers are involved in a remote access, neither on origin or target side: a user level process can directly access remote locations without any kind of software involvement. We have implemented the architecture as an FPGA-based prototype in order to demonstrate the functionality of the complete system. This prototype also allows real world measurements in order to show the performance potential of this architecture, in particular for fine grain memory accesses like they are typically used for synchronization tasks.

We propose a novel class of schemes for the pansharpening of multispectral (MS) images using a multivariate empirical mode decomposition (MEMD) algorithm. MEMD is an extension of the empirical mode decomposition (EMD) algorithm, which enables the decomposition of multivariate data into its intrinsic oscillatory scales. The ability of MEMD to process multichannel data directly by performing data-driven, local, and multiscale analysis makes it a perfect match for pansharpening applications, a task for which standard univariate EMD is ill-equipped due to the nonuniqueness, mode-mixing, and mode-misalignment issues. We show that MEMD overcomes the limitations of standard EMD and yields improved spatial and spectral performance in the context of pansharpening of MS images. The potential of the proposed schemes is further demonstrated through comparative analysis against a number of standard pansharpening algorithms on both simulated Pleiades and real-world IKONOS data sets.

The aging population of China is becoming increasingly more prominent, thus increasing the burden on medical resources. Therefore, the use of data mining technology to improve the efficiency of disease diagnosis has the following important significance. For hospitals, such technology can reduce the cost of providing one-on-one guidance to patients and the probability of registration errors. For patients, it can save time and energy spent on hospital visits; in addition, through remote access, patients can follow the automated guidance at home to complete registration, thereby enhancing admission efficiency. For internet users, such technology enables self-checking of these users’ health conditions on a regular basis; based on certain main symptoms, possible diseases can be pre-diagnosed, thus providing a risk warning. Online medical guidance has become a very important step. To this end, we focus on employing the data mining technology to enhance the performance of online medical guidance. In this paper, we propose a medical diagnosis method called the named entity recognition method and a convolutional neural network model. We apply our proposed method and model as an innovative framework for hospitalization guidance to provide human-like, comprehensive and informative automated medical consultations. We perform experiments on real-world datasets. The experimental results show that our methods achieve state-of-the-art performance compared with baselines.

Measuring airborne fine particulate matter (PM2.5) in the context of health-based field research introduces many logistical challenges including remote equipment setup, data access, and maintenance by both researchers and participants. In preparation for a 17-site, 228-household clinical trial, we developed methods to validate low-cost, light-scattering PM2.5 sensors (model PAII-SD, PurpleAir, Inc, USA) against a Federal Equivalent Method instrument (model BAM 1020, Met One Instruments, USA) for remote measurement of indoor PM2.5 concentrations among participating households over 6 months. To allow real-time, remote data access by the research team, we developed procedures for pairing the light-scattering PM2.5 sensors for each household with a portable Wi-Fi hotspot device (model Solis Lite, Skyroam, Inc, USA) prior to shipping them to participants. Armed with this remote data access, we created a process for daily automated data extractions to ensure sensors remained online and to assess data quality. Our methods support the feasibility of obtaining long-term, continuous PM2.5 measurement reliably and accurately in the context of a research trial using low-cost, light-scattering sensors. The framework we describe in this methods paper can serve as a model for future environmental health research studies.

Recent advances in networking hardware have led to a new generation of Remote Memory Access (RMA) networks in which processors from different machines can communicate directly, bypassing the operating system and allowing higher performance. Researchers and practitioners have proposed libraries and programming models for RMA to enable the development of applications running on these networks, However, the memory models implied by these RMA libraries and languages are often loosely specified, poorly understood, and differ depending on the underlying network architecture and other factors. Hence, it is difficult to precisely reason about the semantics of RMA programs or how changes in the network architecture affect them. We address this problem with the following contributions: (i) a coreRMA language which serves as a common foundation, formalizing the essential characteristics of RMA programming; (ii) complete axiomatic semantics for that language; (iii) integration of our semantics with an existing constraint solver, enabling us to exhaustively generate coreRMA programs (litmus tests) up to a specified bound and check whether the tests satisfy their specification; and (iv) extensive validation of our semantics on real-world RMA systems. We generated and ran 7441 litmus tests using each of the low-level RMA network APIs: DMAPP, VPI Verbs, and Portals 4. Our results confirmed that our model successfully captures behaviors exhibited by these networks. Moreover, we found RMA programs that behave inconsistently with existing documentation, confirmed by network experts. Our work provides an important step towards understanding existing RMA networks, thus influencing the design of future RMA interfaces and hardware.

The Cloud Computing concept has witnessed immense growth, and the rationale behind this increased interest in the emerging paradigm is the cost-effective transmission, storage and intensive computation. The idea is to render remote storage and data analysis capability to the end-user using shared computing resources, thereby bringing down the total cost incurred for an individual. However, consumers are still reluctant to adopt this technology due to associated security and privacy concerns. This manuscript presents a comprehensive description of various threats and technical security concerns with respect to cloud computing. Additionally, support vector machine was used to detect intrusion on the NSL-KDD dataset designed by DARPA. The accuracy of the algorithm for DoS and probe attack were analyzed, and the results are represented in the form of confusion matrices. Among the 41 attributes of the NSL-KDD dataset, only 24 significant attributes are selected. We observed that our model could predict a DoS attack with an accuracy of 96% and a probe attack with an accuracy of 84%.

With the ubiquitous deployment of mobile and sensor technologies in modes of transportation, taxis have become a significant component of public transportation. However, vacant taxis represent an important waste of transportation resources. Forecasting taxi demand within a short time achieves a supply–demand balance and reduces oil emissions. Although earlier studies have forwarded highly developed machine learning- and deep learning-based models to forecast taxicab demands, these models often face significant computational expenses and cannot effectively utilize large-scale trajectory sensor data. To address these challenges, in this paper, we propose a hybrid deep learning-based model for taxi demand prediction. In particular, the Variational Mode Decomposition (VMD) algorithm is integrated along with a Bidirectional Long Short-Term Memory (BiLSTM) model to perform the prediction process. The VMD algorithm is applied to decompose time series-aware traffic features into multiple sub-modes of different frequencies. After that, the BiLSTM method is utilized to predict time series data fed with the relevant demand features. To overcome the limitation of high computational expenses, the designed model is performed on the Spark distributed platform. The performance of the proposed model is tested using a real-world dataset, and it surpasses existing state-of-the-art predictive models in terms of accuracy, efficiency, and distributed performance. These findings provide insights for enhancing the efficiency of passenger search and increasing the profit of taxicabs.