ABSTRACT
The session will report on the success and lessons learned from the five-year implementation of a collaborative DOE project between two Hispanic Serving Institution (HSI) State Colleges and an HSI university with a combined 140,000+ undergraduate students. The session will also report on revising a Systemic, Evidenced-Based, and Student-Centered (SE-SC) framework due to the COVID-19 situation over the last two years. The original aim of the SE-SC framework was to maximize the number of academically-talented, Hispanic students who complete their AA degrees at State Colleges and transfer to a 4-year institution to complete their BS degree and are career-ready to enter engineering and computer science (ECS). The revised SE-SC framework addresses the online education challenge of the project due to the COVID-19 situation. The session will report on how the professional relationships among three large post-secondary institutions have evolved and how the partners have become more intentional about project outcomes. In addition, the design and implementation of articulation agreements have increased programmatic alignment, a more seamless and easy-to-navigate transfer process for students. Furthermore, the collaboration to reach out to industry partners has increased the authenticity of experiences provided to the students across all three institutions. The session will also report on the faculty's adaptation of their instructional practices to include using newer digital technologies for hybrid and remote learning while maximizing student interests and motivating degree completion during the COVID-19 pandemic. Project success has been assessed by applying quantitative and qualitative measures, informal assessments, and anecdotal records. The institutional infrastructure in supporting diverse student interests and success in Electrical Engineering, Computer Engineering, and Computer Science degree programs and their career pathways are presented. Other institutions interested in promoting STEM programs may replicate the implemented model due to its effectiveness, as reported in the session. © American Society for Engineering Education, 2022.
ABSTRACT
The continuous transfer of knowledge within an industrial cluster is a key element for the cluster to maintain its long-term competitiveness. In recent years, various shocks have occurred frequently, and issues related to the resilience of regional economies and industrial clusters have attracted attention. These shocks may also hinder the transfer of knowledge within industrial clusters, thereby limiting the development of industrial clusters. However, most of the previous studies have focused on economic performance, and little attention has been paid to how knowledge exchanges within the firms in cluster are affected by shocks. Therefore, this study adopted a resilience perspective, examined the evolution of innovation modes and proximity within industrial clusters, and constructed a four-quadrant analytical framework consisting of two types of proximity and two types of innovation modes, respectively. We attempted to examine whether shocks cause the shift of innovation modes and proximity in the four-quadrant analytical framework to fill the aforementioned gap. In order to further explored its connotation and make future policy advice more valuable, this study used the concept of life cycle development of clusters to analyze evolution. We took the Hsinchu Science Park as the research object, and used the global financial crisis in 2008 and the Covid-19 in 2020 as the shocks to compare the change of innovation modes and proximity at different life cycle stage. © 2021, Chung Hua University. All rights reserved.
ABSTRACT
The session will report on the success of the last four years of implementing a collaborative DOE project between two state colleges and a Hispanic Serving Institution (HSI) university with a combined 140,000+ undergraduate students. The session will also report the revision of a Systemic, Evidenced-Based, and Student-Centered (SE-SC) framework as initially designed in the DOE project. The original SE-SC framework aims to maximize the number of academically-talented, Hispanic students who complete their AS degrees at State Colleges and transfer to a 4-year institution to complete their B.S. degree and are career-ready to enter engineering and computer science (ECS). The revised SE-SC framework addresses the fully on-line education challenge of undergraduate courses due to the current COVID-19 situation. In particular, the challenges and student outcomes of on-line lab participation are addressed. In addition, the on-line revision of a course-specific mentoring Support Model to ensure student success in completing the Gateway Courses is reported. The overall objective of the mentoring component of the project has been to support students enrolled in gateway mathematics courses to ensure successful course completion. The on-line challenges of mentors and advisors due to the COVID-19 situation are reported in the paper. Data collected for the past four years (2016-2020) validate the proposed initiative's effectiveness. Besides, our innovative approaches to address education, advising, and mentoring challenges due to COVID-19 are presented in the paper. The collaborative model's effectiveness and significance could be replicated among other institutions interested in promoting engineering degrees among Hispanic and low-income students. © American Society for Engineering Education, 2021
ABSTRACT
Over the past two years, large pretrained language models such as BERT have been applied to text ranking problems and showed superior performance on multiple public benchmark data sets. Prior work demonstrated that an ensemble of multiple BERT-based ranking models can not only boost the performance, but also reduce the performance variance. However, an ensemble of models is more costly because it needs computing resource and/or inference time proportional to the number of models. In this paper, we study how to retain the performance of an ensemble of models at the inference cost of a single model by distilling the ensemble into a single BERT-based student ranking model. Specifically, we study different designs of teacher labels, various distillation strategies, as well as multiple distillation losses tailored for ranking problems. We conduct experiments on the MS MARCO passage ranking and the TREC-COVID data set. Our results show that even with these simple distillation techniques, the distilled model can effectively retain the performance gain of the ensemble of multiple models. More interestingly, the performances of distilled models are also more stable than models fine-tuned on original labeled data. The results reveal a promising direction to capitalize on the gains achieved by an ensemble of BERT-based ranking models. © 2021 Owner/Author.
ABSTRACT
Due to the outbreak of the novel coronavirus (or known as COVID-19), people are advised to wear masks when they stay outdoors in many countries. This could result in difficulty for some public safety surveillance systems involving face detection or tracking. Therefore, the development of face detection and tracking algorithms for people wearing face masks is particularly important. In this paper, a real-time tracking algorithm for people with or without face masks is proposed. This algorithm is trained on public face datasets with faces without masks. Although the training does not involve face images of people wearing face masks, we show that the proposed algorithm is robust as it is able to perform well in face tracking for people wearing face masks. We also discuss the possible scenarios where the algorithm could lose track of the target when experimenting in tracking masked faces. This can motivate future research in this area.