教学课程资源库 更多>>

关键词： Data models   Vectors   Anomaly detection   Sampling methods   Training   Computer science   Transformers   Semantics   Predictive models   Feature extraction   Deep learning-based log anomaly detection   data resampling methods   class imbalance   empirical analysis  

摘要： Numerous Deep Learning (DL)-based approaches have gained attention in software Log Anomaly Detection (LAD), yet class imbalance in training data remains a challenge, with anomalies often comprising less than 1% of datasets like Thunderbird. Existing DLLAD methods may underperform in severely imbalanced datasets. Although data resampling has proven effective in other software engineering tasks, it has not been explored in LAD. This study aims to fill this gap by providing an in-depth analysis of the impact of diverse data resampling methods on existing DLLAD approaches from two distinct perspectives. Firstly, we assess the performance of these DLLAD approaches across four datasets with different levels of class imbalance, and we explore the impact of resampling ratios of normal to abnormal data on DLLAD approaches. Secondly, we evaluate the effectiveness of the data resampling methods when utilizing optimal resampling ratios of normal to abnormal data. Our findings indicate that oversampling methods generally outperform undersampling and hybrid sampling methods. Data resampling on raw data yields superior results compared to data resampling in the feature space. These improvements are attributed to the increased attention given to important tokens. By exploring the resampling ratio of normal to abnormal data, we suggest generating more data for minority classes through oversampling while removing less data from majority classes through undersampling. In conclusion, our study provides valuable insights into the intricate relationship between data resampling methods and DLLAD. By addressing the challenge of class imbalance, researchers and practitioners can enhance DLLAD performance.

关键词： Codes   Codecs   Generators   Vectors   Linear codes   Parity check codes   Computer science   Reed-Solomon codes   Telecommunications   Sun   Extended code   MDS code   covering radius   deep hole   GRS code   extended GRS code  

摘要： For a given linear code C of length n over GF(q) and a nonzero vector u in GF(q)(n) , Sun, Ding and Chen defined an extended linear code C(u) of C , which is a generalisation of the classical extended code C(-1) of C and called the second kind of an extended code of C (see Finite Fields Appl., vol. 96, 102401, 2024 and Discrete Math., vol. 347, no. 9, 114080, 2024). They developed some general theory of the extended codes C(u) and studied the extended codes C(u) of several families of linear codes, including cyclic codes, projective two-weight codes, nonbinary Hamming codes, and a family of reversible MDS cyclic codes. The objective of this paper is to investigate the extended codes C(u) of MDS codes C over finite fields. The main result of this paper is that the extended code C(u) of an MDS [n,k] code C remains MDS if and only if the covering radius rho(C-perpendicular to)=k and the vector u is a deep hole of the dual code C-perpendicular to . As applications of this main result, an equivalent statement of MDS Conjecture is presented, the extended codes of the GRS codes and extended GRS codes are investigated, and the covering radii and some deep holes of several families of MDS codes are also determined.

关键词： Target tracking   Cameras   Benchmark testing   Self-supervised learning   Training   Object tracking   Computer science   Multiple object tracking   human association   multi-view cameras   self-supervised learning  

摘要： Multi-view multi-human association and tracking (MvMHAT), is an emerging yet important problem for multi-person scene video surveillance, aiming to track a group of people over time in each view, as well as to identify the same person across different views at the same time, which is different from previous MOT and multi-camera MOT tasks only considering the over-time human tracking. This way, the videos for MvMHAT require more complex annotations while containing more information for self-learning. In this work, we tackle this problem with an end-to-end neural network in a self-supervised learning manner. Specifically, we propose to take advantage of the spatial-temporal self-consistency rationale by considering three properties of reflexivity, symmetry, and transitivity. Besides the reflexivity property that naturally holds, we design the self-supervised learning losses based on the properties of symmetry and transitivity, for both appearance feature learning and assignment matrix optimization, to associate multiple humans over time and across views. Furthermore, to promote the research on MvMHAT, we build two new large-scale benchmarks for the network training and testing of different algorithms. Extensive experiments on the proposed benchmarks verify the effectiveness of our method. We have released the benchmark and code to the public.

关键词： Software As A Service   Computer Science   Software Development Management   Project Management   Computer Science   E Infrastructure   Research And Development   Information Technology   Development Projects   Distribution System   Computer System   High Performance Computing   Digital Networks   Computer Usage   Computing Paradigm   Staff Members   Multidisciplinary Team   Funding Sources   Service Users   Computer Technology   Funding Agencies   Application Programming Interface   Computational Domain   National Science Foundation   Interdisciplinary Knowledge   Virtuous Cycle   Innovation Cycle   File System   Nanoelectronics   Modern Science   Virtualization Technology  

摘要： This article shares the author's experience and views of the creation, development, and operation of software-as-a-service cyberinfrastructures in an academic laboratory. A journey of three successive examples is reviewed: Purdue University Network Computing Hubs (PUNCH), In-Virtual Grid Organizations (In-ViGO) and Integrated Digitized Biocollections (iDigBio). The motivation and impact of each example are described along with their innovative technical characteristics that resulted from translating computer science results and experimental research into useful features. Reflections and recommendations inspired by the author's journey through these examples are discussed, including the need to fund and sustain different lifespans of translational computer science projects and the importance of creating and retaining interdisciplinary human capacity.

关键词： Remaining useful life   Interactive learning   Rolling bearings   Convolution  

摘要： To address the issues of low accuracy, high time cost, and different failure modes in predicting the remaining useful life of rolling bearings, a remaining useful life prediction model based on an interactive learning convolution network with a feature attention mechanism (ILCANet) was proposed in this paper. The model divided time series data into an odd part and an even part, and the interactive learning strategy improved the ability of the model to extract features from long series. The binary tree structure was used to increase the number of network layers and to subdivide sequence features. In the model, residual connection was introduced to prevent the gradient from disappearing. In addition, in order to determine the degree of contribution of different features, a feature attention mechanism was embedded to assign weights to features. Experiments were conducted with PHM2012 and XJTU-SY datasets, and the results showed that the proposed method had better prediction accuracy in RUL prediction than other prediction methods.

关键词： Interdisciplinary teaching   e-learning   Interest   Engagement   Diversity   Gender   Computer science education  

摘要： The field of computer science (CS) is facing a crucial challenge in broadening participation and embracing diversity, especially among underrepresented gender groups. The presented interdisciplinary educational program is an efficient response to this challenge, designed to catalyze diversity in CS through engagement with complex, interest-driven problems. This paper outlines the program's structure, elucidates the pedagogical underpinnings, and reflects on the emergent challenges and opportunities. We delve into how the fusion of CS with other academic disciplines can allure a more varied demographic, emphasizing the engagement of female high school students-a demographic pivotally positioned yet significantly untapped in CS. Through a systematic survey analysis, we measure the program's efficacy in increasing interest in CS and in cultivating an appreciation for its application in addressing real-world, cross-disciplinary challenges. Our findings affirm the program's success in bridging the engagement gap by leveraging students' intrinsic interests, thus charting alternative pathways into the CS field. These insights underscore the critical role of interdisciplinary approaches, establishing anew standard for transformative CS educational methods.

关键词： Energy consumption   Data centers   Optimization   Heuristic algorithms   Energy efficiency   Renewable energy sources   Temperature sensors   Computational modeling   Clouds   Computer science   Cloud data center   energy optimization   reinforcement learning   renewable energy  

摘要： The widespread adoption of cloud data centers has led to a rise in energy consumption, with the associated carbon emissions posing a further threat to the environment. Cloud providers are increasingly moving towards sustainable data centers powered by renewable energy sources (RES). The existing approaches fail to efficiently coordinate IT and cooling resources in such data centers due to the intermittent nature of RES and the complexity of state and action spaces among different devices, resulting in poor holistic energy efficiency. In this paper, a reinforcement learning (RL) based framework is proposed to optimize the holistic energy consumption of sustainable cloud data centers. First, a joint prediction method MTL-LSTM is developed to accurately evaluate both energy consumption and thermal status of each physical machine (PM) under different optimization scenarios to improve the state space information of the RL algorithm. Then, this framework designs a novel energy-aware approach named BayesDDQN, which leverages Bayesian optimization to synchronize the adjustments of VM migration and cooling parameter within the hybrid action space of the Double Deep Q-Network (DDQN) for achieving the holistic energy optimization. Moverover, the pre-cooling technology is integrated to further alleviate hotspot by making full use of RES. Experimental results demonstrate that the proposed RL-based framework achieves an average reduction of 2.83% in holistic energy consumption and 4.74% in brown energy, which also reduces cooling energy consumption by 13.48% with minimal occurrences of hotspots. Furthermore, the proposed MTL-LSTM method reduces the root mean square error (RMSE) of energy consumption and inlet temperature predictions by nearly half compared to LSTM and XGBoost.

关键词： Capital punishment   Sentencing   Aggravating &   mitigating factors  

摘要： Murder trials in which the death penalty is sought involves the jury's consideration of aggravating factors (circumstances necessary for the crime to be death penalty eligible) and mitigating factors (considerations submitted by the defense intended to dissuade a death sentence). Using data from capital murder trials in North Carolina from 1990 to 2023, this study examines the relationship of accepted and the heretofore unexplored impact of rejected aggravating, statutory mitigating, and non-mitigating factors on jury sentencing decisions while controlling for a host of other relevant variables. As expected, accepted aggravating factors statistically increased the odds of a death sentence while accepted statutory mitigating factors decreased those odds. However, in a finding not considered in previous the literature, we also determined that aggravating factors submitted but not accepted by the jury decreased the odds of a death sentence while mitigating factors (both statutory and non-statutory) submitted but not accepted increased the odds. A complex analysis of individual aggravators and mitigators revealed considerable variation in these impacts. The implications for the processing of capital punishment cases are discussed.

关键词： Location awareness   Proposals   Circuits and systems   Annotations   Training   Noise   Computer science   Weakly-supervised learning   temporal action localization   progressive complementary learning  

摘要： Weakly-Supervised Temporal Action Localization (WSTAL) aims to localize and classify action instances in long untrimmed videos with only video-level category labels as supervision. A critical challenge of WSTAL is the large gap between video-level supervision and unavailable snippet-level supervision. Prevailing methods typically assign pseudo labels to snippets, but these methods suffer from significant noise caused by the pseudo snippet-level labels. In this work, we address the WSTAL from a novel category exclusion perspective, which gradually enhances the snippet-level supervision to bridge the gap. Our proposed Progressive Complementary Learning (ProCL) is inspired by the fact that, video-level labels precisely indicate the categories that all snippets surely do not belong to, which is ignored by previous works. Accordingly, we first exclude these surely non-existent categories by the deterministic complementary learning. And then, we introduce the entropy-based pseudo complementary learning that is able to exclude more categories for snippets of less ambiguity. Furthermore, for the remaining ambiguous snippets, we attempt to reduce the ambiguity by distinguishing foreground actions from the background. Extensive experimental results show that our method achieves new state-of-the-art performance on THUMOS14, ActivityNet1.3, and MultiTHUMOS benchmarks.