检索结果-内蒙古大学图书馆

ComPile: A Large IR Dataset from Production Sources

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Grossman, Aiden Paehler, Ludger Parasyris, Konstantinos Ben-Nun, Tal Hegna, Jacob Moses, William S. Monsalve Diaz, Jose M. Trofin, Mircea Doerfert, Johannes University of California Davis United States School of Computation Information and Technology Technical University of Munich Germany Center for Applied Scientific Computing Lawrence Livermore National Laboratory United States University of Minnesota Twin Cities United States Department of Computer Science University of Illinois Urbana-Champaign United States Division of Mathematics and Computer Science Argonne National Laboratory United States Google Inc. United States

Code is increasingly becoming a core data modality of modern machine learning research impacting not only the way we write code with conversational agents like OpenAI’s ChatGPT, Google’s Bard, or Anthropic’s Claude, the way we translate code from one language into another, but also the compiler infrastructure underlying the language. While modeling approaches may vary and representations differ, the targeted tasks often remain the same within the individual classes of models. Yet, relying solely on the ability of modern models to extract information from unstructured code does not take advantage of 70 years of programming language and compiler development by not utilizing the structure inherent to programs in the data collection. This detracts from the performance of models working over a tokenized representation of input code and precludes the use of these models in the compiler itself. To work towards the first intermediate representation (IR) based models, we fully utilize the LLVM compiler infrastructure, shared by a number of languages, to generate a 1.4T Llama 2 token dataset of LLVM IR. We generated this dataset from programming languages built on the shared LLVM infrastructure, including Rust, Swift, Julia, and C/C++, by hooking into LLVM code generation either through the language’s package manager or the compiler directly to extract the dataset of intermediate representations from production grade programs. Statistical analysis proves the utility of our dataset not only for large language model training, but also for the introspection into the code generation process itself as well as for training of machine-learned compiler components. Copyright © 2023, The Authors. All rights reserved.

关键词： Large datasets

CNN-based Classification Framework for Tissues of Lung with Additional Information

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Hu, Huafeng Ye, Ruijie Thiyagalingam, Jeyarajan Coenen, Frans Su, Jionglong Department of Electrical and Electronic Engineering University of Liverpool Based at Xi'an Jiaotong-Liverpool University Jiangsu Suzhou215123 China Department of Computer Science University of Liverpool LiverpoolL69 3BX United Kingdom Scientific Computing Department Science and Technologies Facilities Council Harwell Campus OxfordOX11 0GD United Kingdom School of Ai and Advanced Computing Xi'an Jiaotong-Liverpool University Jiangsu Suzhou215123 China

Interstitial lung diseases are a large group of heterogeneous diseases characterized by different degrees of alveolitis and pulmonary fibrosis. Accurately diagnosing these diseases has significant guiding value for formulating treatment plans. Although previous work has produced impressive results in classifying interstitial lung diseases, there is still room for improving the accuracy of these techniques, mainly to enhance automated decision-making. In order to improve the classification precision, our study proposes a convolutional neural networks-based framework with additional information. Firstly, ILD images are added with their medical information by re-scaling the original image in Hounsfield Units. Secondly, a modified CNN model is used to produce a vector of classification probability for each tissue. Thirdly, location information of the input image, consisting of the occurrence frequencies of different diseases in the CT scans on certain locations, is used to calculate a location weight vector. Finally, the Hadamard product between two vectors is used to produce a decision vector for the prediction. Compared to the state-of-the-art methods, the results using a publicly available ILD database show the potential of predicting these using different additional information. Copyright © 2022, The Authors. All rights reserved.

关键词： Tissue

Laplacian preconditioning of elliptic pdes: Localization of the eigenvalues of the discretized operator

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Gergelits, Tomáš Mardal, Kent-André Nielsen, Bjørn Fredrik Strakoš, Zdeněk Institute of Computer Sciences Czech Academy of Science Prague Czech Republic Faculty of Mathematics and Physics Charles University Prague Czech Republic Department of Mathematics University of Oslo Oslo0316 Norway Department of Numerical Analysis and Scientific Computing Simula Research Laboratory Lysaker1325 Norway Faculty of Science and Technology Norwegian University of Life Sciences ÅsNO-1432 Norway

In the paper Preconditioning by inverting the Laplacian;an analysis of the eigenvalues. IMA Journal of Numerical Analysis 29, 1 (2009), 24-42, Nielsen, Hackbusch and Tveito study the operator generated by using the inverse of the Laplacian as preconditioner for second order elliptic PDEs ∇ · (k(x)∇u) = f. They prove that the range of k(x) is contained in the spectrum of the preconditioned operator, provided that k is continuous. Their rigorous analysis only addresses mappings defined on infinite dimensional spaces, but the numerical experiments in the paper suggest that a similar property holds in the discrete case. Motivated by this investigation, we analyze the eigenvalues of the matrix L−1A, where L and A are the stiffness matrices associated with the Laplace operator and general second order elliptic operators, respectively. Without any assumption about the continuity of k(x), we prove the existence of a one-to-one pairing between the eigenvalues of L−1A and the intervals determined by the images under k(x) of the supports of the FE nodal basis functions. As a consequence, we can show that the nodal values of k(x) yield accurate approximations of the eigenvalues of L−1A. Our theoretical results are illuminated by several numerical experiments.65F08, 65F15, 65N12, 35J99 Copyright © 2018, The Authors. All rights reserved.

关键词： Eigenvalues and eigenfunctions

AClog: Attack Chain Construction Based on Log Correlation

学校读者我要写书评

暂无评论

AClog: Attack Chain Construction Based on Log Correlation

IEEE Global Communications Conference

作者： Teng Li Jianfeng Ma Qingqi Pei Yulong Shen Chi Lin Siqi Ma Mohammad S. Obaidat School of Cyber Engineering Xidian University Shaanxi China Shaanxi Key Laboratory of BlockChain and Security Computing Xidian University Shaanxi China School of Computer Science Xidian University Shaanxi China School of Software Technology Dalian University of Technology Dalian China Commonwealth Scientific and Industrial Research Organisation Sydney Australia College of Computing and Informatics University of Sharjah UAE University of Jordan Jordan Nazarbayev University Kazakhstan University of Science and Technology Beijing China

ISBN: (数字)9781728109626

ISBN: (纸本)9781728109633

Before the final attack happens, clandestine attackers conduct sequenced stages for being stealthy and elusive. These attacks can leave clues in several different log files. However existing approaches can only detect the anomalies using single type of log and fail to reveal all of the attack steps through log integration and correlation. Such methods can hardly detect the relationships among events and prevent the attack in advance. Additionally, traditional machine learning or data mining in log analysis has a high overhead in computing which is impractically applied in a real product or system. To address these problems, we present AClog, a multiple log correlated analysis system to construct the attack chain. Inspired by penetration testing and social network analysis, we transfer the attack provenance as an event relationship discover problem. We use different logs to form the steps of the system and regard them as the event sequences before the attack. Then, we leverage Fast Linear SVM and Longest Common Subsequences to find out the regular steps before the attack. Finally, we spot the corresponding log sequences to identify the pre-attack steps proactively. We apply our approach in the attack prediction of a cloud computing platform and a university network. The results show that the proposed method can effectively and precisely construct the attack steps and identify the corresponding syslogs.

关键词： Correlation Clustering algorithms Support vector machines Windows

Diffusion-based Graph Generative Methods

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Chen, Hongyang Xu, Can Zhen, Lingyu Zhang, Qiang Lin, Xuemin The Research Center for Graph Computing Hangzhou China School of Data Science and Engineering East China Normal University Shanghai China The Research Center for Graph Computing Hangzhou China Hangzhou Institue for Advanced Study UCAS Hangzhou China The College of Computer Science and Technology Zhejiang University Hangzhou China ZJU-Hangzhou Global Scientific and Technological Innovation Center Hangzhou China Antai College of Economics and Management Shanghai Jiao Tong University Shanghai China

Being the most cutting-edge generative methods, diffusion methods have shown great advances in wide generation tasks. Among them, graph generation attracts significant research attention for its broad application in real life. In our survey, we systematically and comprehensively review on diffusion-based graph generative methods. We first make a review on three mainstream paradigms of diffusion methods, which are denoising diffusion probabilistic models, score-based genrative models, and stochastic differential equations. Then we further categorize and introduce the latest applications of diffusion models on graphs. In the end, we point out some limitations of current studies and future directions of future explorations. The summary of existing methods metioned in this survey is in our Github: https://***/zhejiangzhuque/Diffusion-based-Graph-Generative-Methods. Copyright © 2024, The Authors. All rights reserved.

关键词： Generative adversarial networks

学校读者我要写书评

暂无评论

THE RQR ALGORITHM

arXiv 2024年

作者： Camps, Daan Mach, Thomas Vandebril, Raf Watkins, David S. National Energy Research Scientific Computing Center Lawrence Berkeley National Laboratory United States University of Potsdam Institute of Mathematics Karl-Liebknecht-Str. 24–25 Potsdam14476 Germany Department of Computer Science KU Leuven Belgium Department of Mathematics Washington State University United States

Pole-swapping algorithms, generalizations of bulge-chasing algorithms, have been shown to be a viable alternative to the bulge-chasing QZ algorithm for solving the generalized eigenvalue problem for a matrix pencil A − λB. It is natural to try to devise a pole-swapping algorithm that solves the standard eigenvalue problem for a single matrix A. This paper introduces such an algorithm and shows that it is competitive with Francis’s bulge-chasing QR *** Codes 65F15, 15A18 © 2024, CC BY.

关键词： Consensus algorithm

Quadratic pseudospectrum for identifying localized states

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Cerjan, Alexander Loring, Terry A. Vides, Fredy Center for Integrated Nanotechnologies Sandia National Laboratories AlbuquerqueNM87185 United States Department of Mathematics and Statistics University of New Mexico AlbuquerqueNM87123 United States Scientific Computing Innovation Center School of Mathematics and Computer Science Universidad Nacional Autónoma de Honduras Tegucigalpa Honduras

We examine the utility of the quadratic pseudospectrum in photonics and condensed matter. Specifically, the quadratic pseudospectrum represents a method for approaching systems with incompatible observables, as it both minimizes the "eigen-error" in the joint approximate spectrum of the incompatible observables and does not increase the system's computational complexity. Moreover, we derive an important estimate relating the Clifford and quadratic pseudospectra. Finally, we prove that the quadratic pseudospectrum is local, and derive the bounds on the errors that are incurred by truncating the system in the vicinity of where the pseudospectrum is being calculated. Copyright © 2022, The Authors. All rights reserved.

关键词：

Adam-family Methods for Nonsmooth Optimization with Convergence Guarantees

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Xiao, Nachuan Hu, Xiaoyin Liu, Xin Toh, Kim-Chuan Institute of Operations Research and Analytics National University of Singapore 3 Research Link Singapore117602 Singapore School of Computer and Computing Science Hangzhou City University Hangzhou310015 China State Key Laboratory of Scientific and Engineering Computing Academy of Mathematics and Systems Science Chinese Academy of Sciences Beijing100190 China Department of Mathematics Institute of Operations Research and Analytics National University of Singapore 10 Lower Kent Ridge Road Singapore119076 Singapore

In this paper, we present a comprehensive study on the convergence properties of Adam-family methods for nonsmooth optimization, especially in the training of nonsmooth neural networks. We introduce a novel two-timescale framework that adopts a two-timescale updating scheme, and prove its convergence properties under mild assumptions. Our proposed framework encompasses various popular Adam-family methods, providing convergence guarantees for these methods in training nonsmooth neural networks. Furthermore, we develop stochastic subgradient methods that incorporate gradient clipping techniques for training nonsmooth neural networks with heavy-tailed noise. Through our framework, we show that our proposed methods converge even when the evaluation noises are only assumed to be integrable. Extensive numerical experiments demonstrate the high efficiency and robustness of our proposed methods. Copyright © 2023, The Authors. All rights reserved.

关键词： Stochastic systems

Make 1 regularization effective in training sparse CNN

学校读者我要写书评

暂无评论

arXiv 2018年

作者： He, Juncai Jia, Xiaodong Xu, Jinchao Zhang, Lian Zhao, Liang Department of Mathematics Pennsylvania State University University ParkPA16802 United States Department of Computer Science and Engineering Pennsylvania State University University ParkPA16802 United States State Key Laboratory of Scientific and Engineering Computing Academy of Mathematics and Systems Science Chinese Academy of Sciences University of Chinese Academy of Sciences Beijing100190 China

Compressed Sensing using 1 regularization is among the most powerful and popular sparsification technique in many applications, but why has it not been used to obtain sparse deep learning model such as convolutional neural network (CNN)? This paper is aimed to provide an answer to this question and to show how to make it work. Following Xiao (2010), We first demonstrate that the commonly used stochastic gradient decent (SGD) and variants training algorithm is not an appropriate match with 1 regularization and then replace it with a different training algorithm based on a regularized dual averaging (RDA) method. The RDA method of Xiao (2010) was originally designed specifically for convex problem, but with new theoretical insight and algorithmic modifications (using proper initialization and adaptivity), we have made it an effective match with 1 regularization to achieve a state-of-the-art sparsity for the highly non-convex CNN compared to other weight pruning methods without compromising accuracy (achieving 95% sparsity for ResNet-18 on CIFAR-10, for example). Copyright © 2018, The Authors. All rights reserved.

关键词： Stochastic systems