教学课程资源库 更多>>

关键词： Forecasting   Time series analysis   Benchmark testing   Transformers   Predictive models   Data models   Computer science   Reliability   Proposals   Electricity   Benchmarking   multivariate time series   spatial-temporal forecasting   long-term time series forecasting  

摘要： Multivariate Time Series (MTS) analysis is crucial to understanding and managing complex systems, such as traffic and energy systems, and a variety of approaches to MTS forecasting have been proposed recently. However, we often observe inconsistent or seemingly contradictory performance findings across different studies. This hinders our understanding of the merits of different approaches and slows down progress. We address the need for means of assessing MTS forecasting proposals reliably and fairly, in turn enabling better exploitation of MTS as seen in different applications. Specifically, we first propose BasicTS+, a benchmark designed to enable fair, comprehensive, and reproducible comparison of MTS forecasting solutions. BasicTS+ establishes a unified training pipeline and reasonable settings, enabling an unbiased evaluation. Second, we identify the heterogeneity across different MTS as an important consideration and enable classification of MTS based on their temporal and spatial characteristics. Disregarding this heterogeneity is a prime reason for difficulties in selecting the most promising technical directions. Third, we apply BasicTS+ along with rich datasets to assess the capabilities of more than 30 MTS forecasting solutions. This provides readers with an overall picture of the cutting-edge research on MTS forecasting.

关键词： Games   Collaboration   Social networking (online)   Mathematical models   Complexity theory   Roads   Computer science   Vectors   Urban areas   Time factors   Cooperation   non-cooperative game   temporal network   time limitation  

摘要： Temporal networks are obtained from time-dependent interactions among individuals, whereas the interactions can be emails, phone calls, face-to-face meetings, or work collaboration. In this article, a temporal game framework is established, in which interactions among rational individuals are embedded into two-player games in a time-dependent manner. This allows studying the time-dependent complexity and variability of interactions, and the way they affect prosocial behaviors. Based on this simple mathematical model, it is found that the level of cooperation is promoted when the time of collaboration is equally limited for every individual. This observation is confirmed by a series of systematic human experiments on over 1,400 subjects, forming a foundation for comprehensively describing human temporal interactions in collaboration. The research results reveal an important incentive for human cooperation, leading to a better understanding of a fascinating aspect of human nature in society.

关键词： Computer Science   Computer Software   Computer Technology   Areas Of Expertise   Editorial Board   Software Engineering   Relevant Content   Peer Review Process   Technological Domains   Diversity And Inclusion  

关键词： Optimization   Maintenance engineering   Linear programming   Tuning   Computer science   Pareto optimization   Software   Evolutionary computation   Chatbots   Vectors   Adaptive transformation   auxiliary function   constrained optimization   many-objective optimization  

摘要： Transforming a constrained optimization problem (COP) into a multi/many-objective optimization problem (MOP/MaOP) represents a practical approach for solving COPs. This article introduces an adaptive multi/many-objective transformation technique, termed adaptive many-objective transformation technique (AMaOTCO), designed to effectively address COPs. The transformed many-objective optimization problem (MaOP) defines an objective using a convex combination of the objective function (or constraint violation function) and an auxiliary function. This auxiliary function is constructed through a convex combination of the objective function and a weighted constraint violation function. The adaptive tuning of all combination coefficients is based on population information. This adaptive tuning ensures an intelligent balance between minimizing various constraint violations and managing the tradeoff between objective function minimization and constraint violation reduction. The effectiveness of the proposed AMaOTCO is demonstrated through comparisons with state-of-the-art constrained evolutionary algorithms (CEAs) on a set of real-world COPs.

关键词： upward transfer   mixed methods   STEM   computer science   gender  

摘要： Research Questions: While community college transfer (i.e., upward transfer) represents an important mechanism for advancing equity across STEM fields, existing studies of gender and women's participation within computer science have largely excluded the perspectives of upward transfer students. We address this gap in the literature by exploring transfer receptivity and gender discrimination within computer science, guided by the following questions: (1) How do upward transfer computer science students report their receptivity experiences, and how might this differ by gender? (2) How do upward transfer computer science students make meaning of receptivity experiences, and how might that meaning making be shaped by gender? Methods: We use a sequential mixed methods design, relying on longitudinal survey and interview data from upward transfer computer science majors, collected throughout students' first year at the receiving university. Results: Findings reveal that, relative to men, upward transfer women report greater experiences of transfer stigma and challenges accessing resources at the receiving university. Qualitative findings document additional nuances in how upward transfer students-especially women-describe resilience as they navigate the university campus, encounter navigation challenges at the university, and make meaning of various manifestations of transfer stigma on campus. Contributions: In addition to implications for research and theory, we discuss what universities can do to foster a more receptive environment for upward transfer women. Specific recommendations focus on ensuring that spaces for women in computing are inclusive of transfer students and, likewise, creating supportive transfer cohort communities that are inclusive of women.

关键词： 科学合作   机构合作网络   合作绩效   学术影响力  

摘要： [目的/意义]在科学合作中，机构是科学研究中的主要驱动单位。相较机构内合作，跨机构合作往往更具产出高影响力文章的潜力。因此，对机构层面的细粒度合作进行研究具有重要意义。[方法/过程]为了研究不同机构的合作类型与学术影响力之间的关系，本文通过对机构进行分类，定义不同的类型，运用网络分析法，研究不同机构合作类型网络指标与学术影响力之间的关系。[结果/结论]本文以计算机科学领域为例，通过对计算机科学领域不同的机构合作类型网络指标与学术影响力之间的关系分析发现，度中心性和接近中心性与学术影响力呈现正相关的关系，中介中心性则与学术影响力呈负相关关系；而在不同的子领域中，不同的中心性指标与学术影响力之间存在较大差异，与计算机科学领域的整体回归结果也存在较大差异。

关键词： Context modeling   Transformers   Translation   Decoding   Vectors   Training   Neural machine translation   Standards   Computer science   Computational modeling   Machine translation   document embeddings  

摘要： Standard neural machine translation (NMT) assumes that document-level context information is irrespective. Most existing document-level NMT methods are satisfied with a smattering sense of shallow document-level information, such as using a few context sentences surrounding the source sentence as document-level information. Our work focuses on exploiting detailed document-level context in terms of multiple forms of document embeddings, which can sufficiently model deeper and richer document-level context. The proposed document-level NMT is implemented to enhance the Transformer baseline by introducing both global and local document-level clues on the source end. We compress the entire document text with explicit boundaries into a token-size global static document embedding, and the neighboring sentences as a token-size local dynamic document embedding and concatenate with the source tokens. Experiments reveal that the proposed method significantly improves the translation performance over strong baselines and other related studies.

关键词： Activity recognition   Long short term memory   Feature extraction   Context modeling   Computer science   Attention mechanisms   Deep learning   Graph convolutional networks   Computer vision   Adaptation models   Group activity recognition   actor relation graph   spatio-temporal feature extraction   GCN-LSTM-Attention  

摘要： The challenge of understanding and recognizing group activities through human behavior and interactions is a prominent issue in the realm of computer vision research. This area boasts a wide range of applications, including security surveillance, healthcare monitoring, and human-computer interaction systems. However, accurately deciphering complex group dynamics continues to pose significant difficulties. Traditional video-based activity recognition methods often grapple with persistent obstacles such as environmental noise, background clutter, and the inter-class similarities inherent in activity patterns. Recent advancements in this field have largely focused on spatial feature extraction, which alone proves inadequate for thorough group activity analysis. In response to these limitations, we propose a novel deep learning framework that effectively captures both temporal and spatial characteristics of group interactions. Our architecture employs a Convolutional Neural Network (CNN) with Inception-V3 as the foundational model for initial feature extraction. This is complemented by the development of an Actor Relation Graph (ARG) utilizing Zero Normalized Cross Correlation (ZNCC), which adeptly illustrates both appearance-based and positional relationships among participants. By integrating the ARG with a hybrid model that combines Graph Convolutional Network (GCN), Long Short-Term Memory (LSTM), and Attention mechanisms, our approach significantly enhances the extraction of spatial and relational features compared to conventional techniques. Experimental evaluations conducted on two benchmark datasets-the Collective Activity Dataset (CAD) and the Volleyball dataset-demonstrate the efficacy of our framework. Our proposed model achieves state-of-the-art performance, attaining prediction accuracies of 94.32% and 94.47% on the CAD and Volleyball datasets, respectively, thereby surpassing existing methodologies in the realm of group activity recognition.

关键词： User experience   Usability   Open source software   Guidelines   Biological system modeling   Standards organizations   Companies   Computer science   Computational modeling   User centered design   Exploratory factor analysis   free and open-source software   FOSS UX enhancement model   Kaiser Meyer Olkin   principles axis factoring   user experience   user experience maturity influencing factor  

摘要： Free and open-source software has rapidly grown and garnered considerable attention in recent years. However, the design complexities and guidelines inherent to Free and Open Source Software (FOSS) pose a challenging task in enhancing the User Experience (UX) of FOSS. Several pre-existing models have significantly contributed to the existing body of knowledge, yet they often overlook FOSS-specific attributes that establish a direct correlation with the efficiency of the UX in FOSS. To address this gap, the study proposes a FOSS UX enhancement model (FUEM) designed to align with current FOSS practices and guidelines, aiming to improve user experience in free and open-source software. The proposed model consists of six levels, with each level encompassing UX maturity influencing factors (UMXIF) identified in prior research. These factors are re-evaluated through a Likert scale questionnaire to gather feedback from developers, project managers, and UI/UX designers, as the previous study solely included the perspectives of UX experts. Our findings reveal that these factors positively impact FOSS projects, and integrating them into FUEM will enable the FOSS community to ascertain their level of UX maturity and identify which missing UMXIF they must adopt to progress to the next stage of UX maturity in FUEM. Furthermore, the tailored model developed for FOSS is validated through feedback analysis of expert reviews and quantitative analysis employing a one-way ANOVA test. The ANOVA test indicates that the p-value exceeds the significant threshold of. 05, suggesting that the FOSS UX model can considerably enhance UX in FOSS projects by providing a structured pathway for improving UX maturity, serving as a valuable resource for the FOSS community, guiding the development of more user-centric open-source software, and fostering a deeper understanding of UX maturity practices explicitly tailored for FOSS environments.

关键词： IP networks   Control systems   Protocols   Software   Decision trees   Classification algorithms   Memory management   Software defined networking   Field programmable gate arrays   Computer science   Packet classification   hash table   tuple space search   linear search   OpenFlow  

摘要： Packet classification that attracts wide interests in router designs is one of the most important functionalities to provide services such as QoS, packet forwarding, firewall, and VPN. With the expanding network and the rise of software-defined networks, software routers/switches are needed to enable a wide variety of network applications in physical or virtual networks. Existing packet classification schemes used in software switches had become difficult to support a fast classification process for large multi-field rulesets. In addition, it is necessary to have fast updates that are frequently performed. In this paper, we propose a hash-based partitioning scheme for a multi-field packet classification problem. The proposed design is a 4-level data structure based on hashing. It selects the suitable fields to construct the data structures so that we can quickly narrow down the search space to find matched rules. The hash table used in the proposed design is more effective than the existing search data structures. Our experimental results show that the proposed scheme performs best in all four-performance metrics of memory consumption, construction time, classification time, and update time compared with the six most recently proposed schemes, CutTSS, TabTree, TupleMerge, CutSplit, PartitionSort, and PSTSS. The average number of bytes required per rule is only 27-99% of that needed in these schemes. The construction time of our scheme is 2.06 to 327.82 times faster than these schemes. Most importantly, the proposed scheme is 1.92-16.65 and 2.22-4.54 times faster than these six schemes in throughput and update time, respectively.