教学课程资源库 更多>>

关键词： Entropy   Feature extraction   Reliability   Adaptation models   Training   Cyberspace   Computer science   Transfer learning   partial domain adaptation   deep neural network   image classification   contrastive learning  

摘要： Partial domain adaptation (PDA) assumes that target domain class label set is a subset of that of source domain, while this problem setting is close to the actual scenario. At present, there are mainly two methods to solve the overfitting of source domain in PDA, namely the entropy minimization and the weighted self-training. However, the entropy minimization method may make the distribution prediction sharp but inaccurate for samples with relatively average prediction distribution, and cause the model to learn more error information. While the weighted self-training method will introduce erroneous noise information in the self-training process due to the existence of noise weights. Therefore, we address these issues in our work and propose self-training contrastive partial domain adaptation method (STCPDA). We present two modules to mine domain information in STCPDA. We first design self-training module based on simple samples in target domain to address the overfitting to source domain. We divide the target domain samples into simple samples with high reliability and difficult samples with low reliability, and the pseudo-labels of simple samples are selected for self-training learning. Then we construct the contrastive learning module for source and target domains. We embed contrastive learning into feature space of the two domains. By this contrastive learning module, we can fully explore the hidden information in all domain samples and make the class boundary more salient. Many experimental results on five datasets show the effectiveness and excellent classification performance of our method.

关键词： crime   criminal behavior   violent behavior   ACEs   capital punishment   youthful offenders  

摘要： In the 2005 Roper v. Simmons decision, the Supreme Court determined that capital punishment, the death penalty, was not appropriate for offenders under the age of 18. The Court accepted the argument that juveniles are less blameworthy than adult offenders due to differences in their cognitive abilities and the influence of factors outside of their control, such as household dysfunction or the existence of adverse childhood experiences (ACEs). Recent empirical literature examining both brain development and the impact of ACEs on both brain maturity and offending behaviors demonstrates that brain maturation and psychosocial maturation continue into the mid-twenties. As a result, youthful offenders, those 18 to 24 of age, may possess many of the neuropsychological deficiencies attributed to juveniles and yet be sentenced to death. This research examines if juries of capital punishment trials recognize the impact of age and ACEs when determining their sentencing decision.

关键词： allied field   computer science   sociology   cognition   collective intelligence   culture   diversity   formal models   intragroup processes   networks   social cognition  

摘要： Humans regularly solve complex problems in cooperative teams. A wide range of mechanisms have been identified that improve the quality of solutions achieved by those teams on reaching consensus. We argue that many of these mechanisms work via increasing the transient diversity of solutions while the group attempts to reach a consensus. These mechanisms can operate at the level of individual psychology (e.g., behavioral inertia), interpersonal communication (e.g., transmission noise), or group structure (e.g., sparse social networks). Transient diversity can be increased by widening the search space of possible solutions or by slowing the diffusion of information and delaying consensus. All of these mechanisms increase the quality of the solution at the cost of increased time to reach it. We review specific mechanisms that facilitate transient diversity and synthesize evidence from both empirical studies and diverse formal models-including multiarmed bandits, NK landscapes, cumulative-innovation models, and evolutionary-transmission models. Apparent exceptions to this principle occur primarily when problems are sufficiently simple that they can be solved by mere trial and error or when the incentives of team members are insufficiently aligned. This work has implications for our understanding of collective intelligence, problem solving, innovation, and cumulative cultural evolution.

关键词： O10   O32   Digital financial inclusion   Economic growth   Financial development   Technological progress  

摘要： This study employs the panel vector autoregression model to investigate how digital financial inclusion (DFI), technological advancement, environmental conditions, and economic growth are interconnected in 16 low financial development countries (LFDCs) and 29 high financial development countries (HFDCs) from 2015 to 2020. The results of the impulse response function reveal that in LFDCs, DFI enhances environmental quality while promoting technological progress. However, this improvement in environmental quality results in a decline in economic growth in these countries. In HFDCs, the promotion of DFI results in economic growth;however, it is accompanied by a decrease in environmental quality. Furthermore, the results of variance decomposition demonstrate that the interconnection among DFI, economic growth, environmental quality, and technological progress is more closely related in LFDCs than in HFDCs. Based on the findings, we recommend relevant policy implications for the respective country groups.

关键词： Ethics   literature review   computing education   computer science education  

摘要： The computing education research community now has at least 40 years of published research on teaching ethics in higher education. To examine the state of our field, we present a systematic literature review of papers in the Association for Computing Machinery computing education venues that describe teaching ethics in higher-education computing courses. Our review spans all papers published to SIGCSE, ICER, ITiCSE, CompEd, Koli Calling, and TOCE venues through 2022, with 100 papers fulfilling our inclusion criteria. Overall, we found a wide variety in content, teaching strategies, challenges, and recommendations. The majority of the papers did not articulate a conception of "ethics," and those that did used many different conceptions, from broadly applicable ethical theories to social impact to specific computing application areas (e.g., data privacy and hacking). Instructors used many different pedagogical strategies (e.g., discussions, lectures, assignments) and formats (e.g., stand-alone courses, incorporated within a technical course). Many papers identified measuring student knowledge as a particular challenge, and 59% of papers included mention of assessments or grading. Of the 69% of papers that evaluated their ethics instruction, most used student self-report surveys, course evaluations, and instructor reflections. While many papers included calls for more ethics content in computing, specific recommendations were rarely broadly applicable, preventing a synthesis of guidelines. To continue building on the last 40 years of research and move toward a set of best practices for teaching ethics in computing, our community should delineate our varied conceptions of ethics, examine which teaching strategies are best suited for each, and explore how to measure student learning.

关键词： Task analysis   Crowdsourcing   Heuristic algorithms   Computer science   Approximation algorithms   Sun   Schedules   Opportunistic networks   ONs   mobile crowdsourcing   task assignment   mobile social network   priority-aware  

摘要： Mobile crowdsourcing technology is a special type of crowdsourcing that outsources location-based human tasks to workers with mobile devices. The user recruitment and task assignment strategies are critically important for successful mobile crowdsourcing. A number of task assignment algorithms for opportunistic network-based mobile crowdsourcing have been proposed for different design goals. However, none of them considers the priority of tasks. In this article, we introduce the priority-aware task assignment that maximizes the successful task rate of prioritized tasks by keeping with the overall makespan of the tasks as short as possible. To this end, the priority-aware offline task assignment (PFTA) and online task assignment (PNTA) are proposed. Since the task assignment problem is NP-hard, the proposed PFTA and PNTA rely on the greedy strategy based on the priority of tasks and/or the workload of tasks. In addition, the approximation ratio of PFTA and the competitive ratio of PNTA are analyzed. The performance of the proposed schemes are evaluated using real mobility traces, and the simulation results demonstrate that PFTA and PNTA achieve their design goals.

关键词： Program comprehension   source code   C plus   Python   bug fixing   new feature tasks   programming education   learning behavior   eye-tracking study  

摘要： Background and Context: Understanding how a student programmer solves different task types in different programming languages is essential to understanding how we can further improve teaching tools to support students to be industry-ready when they graduate. It also provides insight into students' thought processes in different task types and languages. Few (if any) studies investigate whether any differences exist between the reading and navigation behavior while completing different types of tasks in different programming languages. Objectives: We investigate whether the use of a certain programming language (C++ versus Python) and type of task (new feature versus bug fixing) has an impact on performance and eye movement behavior in students exposed to both languages and task types. Participants: Fourteen students were recruited from a Python course that taught Python as an introductory programming language. Study Method: An eye tracker was used to track how student programmers navigate and view source code in different programming languages for different types of tasks. The students worked in the Geany Integrated Development Environment (IDE, used also in their course) while eye-tracking data was collected behind the scenes making their working environment realistic compared to prior studies. Each task type had a Python and C++ version, albeit on different problems to avoid learning effects. Standard eye-tracking metrics of fixation count and fixation durations were calculated on various areas of the screen and on source code lines. Normalized versions of these metrics were used to compare across languages and tasks. Findings: We found that the participants had significantly longer average fixation duration and total fixation duration adjusted for source code length during bug fixing tasks than the feature addition tasks, indicating bug fixing is harder. Furthermore, participants looked at lines adjacent to the line containing the bug more often before looking at

摘要： Students' dropout of Computer Science (CS) education is a crucial issue. This study aims to investigate one of the aspects which can help to retain our students. It is vital to examine CS education on the challenge of competence transition within the BSc curriculum from faculty perspectives. Teachers' expectations about students' skills and knowledge are important to understand because they influence learning outcomes and teaching methodologies. Acquiring hard skills and professional skills has outstanding significance in preparing students for their future careers. This study uses Q-methodology to identify the different viewpoints on the skills necessary to obtain a CS degree. Teachers from CS bachelor's programmes at a large European university participated in the study and shared their opinions. The participants were asked to rank the statements along a spectrum of "the most important skill" to "unimportant skill" (containing hard skills as well as soft skills). Factor analysis revealed five factors that are key components to obtaining a degree in CS: 1. analytical and technical skills, 2. teamwork and self-study experience, 3. group programming experience, 4. communication and problem-solving skills, 5. mathematical foundations and process modelling ability. This exploratory study applied a new research instrument and approach to incorporate teachers' perspectives into research and practice. These findings could help administrators develop new curricula in order to increase students' retention. We confirmed the need for acquiring professional skills and highlighted the need for designing new programmes which can improve students' soft skills to prepare them for work in the IT field.

关键词： Quantum information   Computer science   Quantum mechanics  

摘要： Finding the ground state of a quantum many-body system is a fundamental problem in quantum physics. In this work, we give a classical machine learning (ML) algorithm for predicting ground state properties with an inductive bias encoding geometric locality. The proposed ML model can efficiently predict ground state properties of an n-qubit gapped local Hamiltonian after learning from only O(log(n)) data about other Hamiltonians in the same quantum phase of matter. This improves substantially upon previous results that require O(nc) data for a large constant c. Furthermore, the training and prediction time of the proposed ML model scale as O(nlogn) in the number of qubits n. Numerical experiments on physical systems with up to 45 qubits confirm the favorable scaling in predicting ground state properties using a small training *** work proposed a machine learning algorithm for predicting ground state properties of quantum many-body systems that outperforms any non-learning classical algorithm but requires extensive training data. Lewis et al. present an improved algorithm with exponentially reduced training data requirements.