February 2024 - Gradstudents - mailman.cs.vt.edu

Jie Bu's Ph.D. defense [Feb 9 at 4 pm]
by Anuj Karpatne 02 Dec '24

02 Dec '24

Hello everyone, You are cordially invited to the final defense of PhD student Jie Bu on Thursday, Feb 9 at 4 pm. Zoom Link: https://virginiatech.zoom.us/j/5142223333 Title: Achieving More with Less: Learning Generalizable Neural Networks With Less Labeled Data and Computational Overheads Committee: Dr. Anuj Karpatne (advisor) Dr. Adrian Sandu Dr. Ismini Lourentzou Dr. Naren Ramakrishnan Dr. Anish Arora (from Ohio State University) Abstract: Recent advancements in deep learning have demonstrated its incredible ability to learn generalizable patterns and relationships automatically from data in a number of mainstream applications. However, the generalization power of deep learning methods largely comes at the costs of working with very large datasets and using highly compute-intensive models. Many applications cannot afford these costs needed to ensure generalizability of deep learning models. For instance, obtaining labeled data can be costly in scientific applications, and using large models may not be feasible in resource-constrained environments involving portable devices. This thesis aims to improve efficiency in machine learning by exploring different ways to learn generalizable neural networks that require less labeled data and computational resources. We demonstrate that using physics supervision in scientific problems can reduce the need for labeled data, thereby improving data efficiency without compromising model generalizability. Additionally, we investigate the potential of transfer learning powered by transformers in scientific applications as a promising direction for further improving data efficiency through domain adaptation. On the computational efficiency side, we present two efforts for increasing parameter efficiency of neural networks through novel architectures and structured network pruning. Thanks, -Anuj ------------------- Anuj Karpatne Assistant Professor Computer Science, Virginia Tech https://people.cs.vt.edu/karpatne/

2 6

Muskan Gupta's Thesis exam
by Sang Won Lee 02 May '24

02 May '24

Hi, You are cordially invited to Muskan Gupta's thesis exam. I've included the details below. Sang Won Lee *When/Where* 8/31/2023 (Thursday) 1 PM-3 PM The Gilbert Place Room 3211 *Committee* Sang Won Lee (Chair) Eugenia Rho Scott McCrickard *Title*: Understanding Social Media Users' Perceptions of Trigger and Content Warnings *Abstract*: Exposure to distressing content on social media poses risks to users' mental well-being, leading to the use of trigger warnings (TW) and content warnings (CW). However, there is limited insight into user preferences and the factors affecting these warnings' effectiveness. This study addressed this gap through qualitative interviews with 15 participants. Findings revealed generally positive user perceptions of warnings, which provide protection from distressing content, create safe spaces for sharing sensitive experiences, and avoid such content in public spaces. However, understanding when, what, and how to use warnings varied among users, leading to complex decision-making. Factors influencing warning effectiveness included their separation from content, length, awareness, consistency, context, severity, reasoning, creator's intent, and audience knowledge. Participants stressed the need for enhanced education on warnings and triggers in social media. The study produced design recommendations, including personalized filtering, platform guidelines, specificity, nudging creators, interventions against excessive scrolling, accessibility improvements, and community feedback mechanisms. Participants also suggested post-trigger support measures. The study uncovers challenges in balancing warnings with user engagement, specificity, and personalization while avoiding the creation of filter bubbles. While the study spanned various platforms, future research should focus on specific platforms to address limitations like recruitment bias and broad scope. Future directions include incentivizing warning use, implementing nudges and interventions, and improving post-trigger support, as well as investigating audio triggers and quantitative studies for refined TW/CW design to promote a trauma-informed social media environment.

2 2

CS Grad Council Meeting -- Thursday 2/29 11am to 12 noon in Gilbert 4001
by Manesh, Daniel 28 Feb '24

28 Feb '24

Hi everyone! We will be holding our next CS Grad Council meeting on Thursday, February 29th from 11am to noon in Gilbert 4001 (or on Zoom<https://virginiatech.zoom.us/j/4995854459>). The agenda items are similar to last week: planning the upcoming graduate recruitment weekend, planning more social events, and discussing a climate survey for the department. Hope to see folks there! Danny Manesh CS Grad Council President

1 0

Scholarships available for CAPWIC 2024!
by Mohammed Seyam 28 Feb '24

28 Feb '24

Greetings! CAPWIC is the ACM Capital Region Celebration of Women in Computing and provides a low-cost, regionally tailored, small conference for women and minorities in computing. The participants include women (students, faculty, and professionals), as well as all supporters of women in computing. The conference will be hosted by *University of Virginia, on April 5-6, 2024*. All levels of experience are welcome. For more information on the conference, visit https://capwic.org/ The Computer Science Department at Virginia Tech is offering a number of *fully-funded scholarships* for interested students to attend the conference. If you are interested in attending the conference, please *apply *through this link <https://bit.ly/CAPWIC2024CSVT>: https://bit.ly/CAPWIC2024CSVT . *Deadline for submission is March 4, 2024*. Applicants who will be offered scholarships will be notified via email. Please reach out to me (seyam(a)vt.edu) if you have any questions. -- *Mohammed Seyam <https://website.cs.vt.edu/people/faculty/mohammed-seyam.html>, *PhD Collegiate Assistant Professor Coordinator for Experiential Learning <https://xl.vt.edu/> Department of Computer Science <http://cs.vt.edu/> Virginia Polytechnic Institute and State University (VirginiaTech <http://www.vt.edu/>). Website <https://cs.vt.edu/people/faculty/mohammed-seyam.html>|LinkedIn <https://www.linkedin.com/in/msseyam/> | Facebook <https://www.facebook.com/MSSeyam>

1 0

Join us for a CS Grad Alumni Industry Panel
by Sara Hooshangi 27 Feb '24

27 Feb '24

The CS-MEng team invites you to join us on Friday, March 1, 4:00 – 5:00 PM on Zoom for a CS Grad Alumni Industry Panel Discussion. Hear from alumni and professionals about their day-to-day work, trends in the field and how to best position yourself with skills and experiences to be competitive for full-time opportunities. Here are the panelists. Hannah Chen currently works as a Software Engineer at Bloomberg in New York City. Hannah graduated from the MEng program in Fall 2023 with a concentration in Software Engineering. Jordan Gillard currently works as a Software Engineer at Bloomberg in New York City. Jordan graduated from the MEng program in Spring 2021 with a concentration in Software Engineering. Luisa Guerra-Young currently works as a Senior Executive Communicator at The Boeing Company in Arlington, Virginia. Luisa is currently is pursuing her MEng degree part-time and began the program in Fall 2023. Dvijen Trivedi currently works as a Software Engineer at Amazon in Seattle, Washington. Dvijen graduated from the MEng program in Fall 2022 with a concentration in AI and Machine Learning. Topic: MEng CS Industry Panel Time: Mar 1, 2024 04:00 PM Eastern Time (US and Canada) Join Zoom Meeting https://virginiatech.zoom.us/j/3888174704?pwd=SGRGTkFFRVozcUdZYjA1SkRPaTllQ… <https://virginiatech.zoom.us/j/3888174704?pwd=SGRGTkFFRVozcUdZYjA1SkRPaTllQ…> Meeting ID: 388 817 4704 Passcode: 98482 ---------------------------------- Sara Hooshangi, Ph.D. Collegiate Associate Professor Director, MEng Program Department of Computer Science Virginia Tech

1 0

REMINDER! Giving Day and PANCAKES!!
by Watson, Joan 22 Feb '24

22 Feb '24

Good morning, everyone! There are 2½ hours left of GIVING DAY<https://cs.vt.edu/>, so make your gift count! ... AND don’t forget to stop by 1120 Torgersen Hall for fresh-off-the-griddle PANCAKES prepared by Dr. Cal Ribbens, CS Department Head. He’ll be cooking until 12:00 PM today, so “Come and get ‘em!” Your donation to the department goes a long way in supporting enriching and beneficial programming for students, including scholarships, funding to attend conferences, and opportunities to engage in exciting capstone and research projects. Make your gift HERE<https://givingday.vt.edu/giving-day/81643/donate>. Thank you for giving. Thank you for serving. UT PROSIM *************** Joan Monahan Watson, Ph.D. Director of Administrative Operations Department of Computer Science Virginia Tech 3160D Torgersen Hall 620 Drillfield Drive Blacksburg, VA 24061 Phone: 540.231.8874 Email: jmwatson(a)vt.edu<mailto:jmwatson@vt.edu>

1 0

Survey on upcoming HCI course(s) in the CS Department -- Due March 1
by Williams, Ihudiya 22 Feb '24

22 Feb '24

Hi Everyone, I wanted to take a human-centered approach for my upcoming seminar course and future courses. I created this survey for students to provide feedback on what HCI course(s) they would like to have within the computer science department. Please complete this short survey by Friday, March 1. Link: https://forms.gle/NssFtkrcJTLFcbdv9 Please note this survey is optional but super helpful. Best, Dr. Williams -- [https://ci3.googleusercontent.com/mail-sig/AIorK4zDEVb9qkqBTdIIH2dlXN0GcICb…] Ihudiya Finda Williams, Ph.D. Assistant Professor Virginia Tech | Department of Computer Science Address: 620 Drillfield Drive, Blacksburg, VA 24061 Email: iogburu(a)vt.edu<mailto:iogburu@umich.edu> Website<http://www.ihudiyawilliams.com/> | LinkedIn<http://www.linkedin.com/in/ihudiyawilliams> | Schedule a Meeting<https://outlook.office.com/bookwithme/user/c5efbccff6f543b9b9401e7adf09e322…> My work hours may not align with yours. Please feel no obligation to respond outside of your office time. Virginia Tech acknowledges that we live and work on the Tutelo / Monacan People’s homeland, and we recognize their continued relationships with their lands and waterways. We further acknowledge that the Morrill Land-Grant College Act (1862) enabled the Commonwealth of Virginia to finance and found Virginia Tech through the forced removal of Native Nations from their lands in western territories. We understand that honoring Native Peoples without explicit material commitments falls short of our institutional responsibilities. Through sustained, transparent, and meaningful engagement with the Tutelo / Monacan Peoples, and other Native Nations, we commit to changing the trajectory of Virginia Tech's history by increasing Indigenous student, staff, and faculty recruitment and retention, diversifying course offerings, and meeting the growing needs of all Virginia tribes and supporting their sovereignty. Virginia Tech acknowledges that its Blacksburg campus sits partly on land that was previously the site of the Smithfield and Solitude Plantations, owned by members of the Preston family. Between the 1770s and the 1860s, the Prestons and other local White families that owned parcels of what became Virginia Tech also owned hundreds of enslaved people. We acknowledge that enslaved Black people generated wealth that financed the predecessor institution to Virginia Tech, the Preston and Olin Institute, and they also worked on the construction of its building. Not until 1953, however, was the first Black student permitted to enroll. Through InclusiveVT, the institutional and individual commitment to Ut Prosim (that I may serve) in the spirit of community, diversity, and excellence, we commit to advancing a more diverse, equitable, and inclusive community. Privacy Notification: This message (including any attachments) is intended only for the use of the individual or entity to which it is addressed and may contain information that is non-public, proprietary, privileged, confidential, and exempt from disclosure. If you are not the intended recipient, you are hereby notified that any use, dissemination, distribution, or copying of this communication is strictly prohibited. If you have received this communication in error, destroy this message if it is a facsimile, or (ii) delete this message immediately if this is an electronic communication.

1 0

Bowen Shen's PhD Prelim exam scheduled at 8:30am on Thursday 02/22/24
by Na Meng 21 Feb '24

21 Feb '24

Hi all, You're warmly welcomed to attend Bowen's PhD prelim exam. *Title:* Understanding and Automatically Detecting Software Merge Conflicts. *Date: *Thursday, Feb 22nd, 2024 *Time:* 8:30am-10:30am EST *Location:* Gilbert Place, room 4307 *Zoom link:* https://virginiatech.zoom.us/j/86527390066 *Abstract: *In collaborative software development, developers create branches to add features and fix bugs tentatively to work in parallel. Periodically developers need to merge the branches to integrate edits. However, merge conflicts can arise when edits overlap or interfere with each other. Resolving merge conflicts is challenging and tedious. Prior work proposed tools to detect and solve conflicts. However, many questions are still not fully investigated, such as what types of conflicts exist in real-world applications and how developers or tools handle them. To solve this problem, we conducted an empirical study to investigate merge conflicts in 208 open-source repositories. We manually inspected 538 conflicts with three types (textual, build and test) and their resolutions to characterize merge conflicts from different angles. We also designed a benchmark, ConflictBench, to test various tools in terms of their capabilities of conflict detection and resolution. ConflictBench consists of 180 merging scenarios extracted from 180 open-source Java projects. We defined three novel metrics to evaluate merge tools. By applying five state-of-the-art tools to ConflictBench, we observed that ConflictBench is effective to characterize different tools. It helps reveal limitations of existing tools and sheds light on future research. Lastly, to help developers detect testing conflicts that are not well detected by existing tools, we will design a new approach to automatically generate test cases to reveal potential conflicts. This tool will generate test cases targeting merge-related code regions, via genetic programming and program static analysis. In summary, our research characterizes real-world merge conflicts as well as software merge tools, and proposes a new approach to better detect conflicts. Our work will help developers better tackle the technical challenges posed by software merge conflicts, and shed light on future research in that area. Cheers, Na -- *Na Meng, Associate Professor* Virginia Tech | Computer Science (0902) Gilbert Place, Suite 4311A 220 Gilbert Street, Blacksburg, VA 24060 (540) 231-5492 | nm8247(a)vt.edu

1 0

M.S. Thesis Defense: A 3D Deep Learning Architecture for Denoising Low-Dose CT Scans
by Wuchun Feng 20 Feb '24

20 Feb '24

You are cordially invited to the M.S. final defense of Armen Kasparian (ECE). When: Wednesday, February 21 @ 11:00 am Where: Torgersen 2050 *OR* https://virginiatech.zoom.us/j/5928594805?omn=88648764340 Title: A 3D Deep Learning Architecture for Denoising Low-Dose CT Scans Abstract: Low-dose computed tomography (LDCT) scans reduce the radiation dose of computed tomography (CT) scans but come at the expense of image quality. Deep-learning (DL) image denoising techniques can enhance these LDCT images to match the quality of their regular-dose CT counterparts. To achieve better denoising performance than the current state of the art, we present a novel 3D DL architecture for LDCT image denoising called 3D-DDnet. The architecture leverages the interslice correlation in volumetric CT scans to obtain better denoising performance and employs distributed data parallel (DDP) strategies along with transfer learning to achieve faster training. The DDP training strategy enables a scalable multi-GPU approach on Nvidia A100 GPUs, which allows the training of previously prohibitively large volumetric samples. Our results show that 3D-DDnet achieves 10% better mean square error (MSE) on LDCT scans than its 2D predecessor (i.e., 2D-DDnet). In addition, the transfer learning in 3D-DDnet leverages existing trained 2D models to “jump start” the weights and biases of our 3D DL model and reduces training time by 50% while maintaining accuracy Join Zoom Meeting https://virginiatech.zoom.us/j/5928594805?omn=88648764340 Meeting ID: 592 859 4805 ~~ Wu FENG | Synergy Lab | Depts. of CS, ECE, BEAM | NSF SHREC Center & SEEC Center at Virginia Tech | Professor, Turner Faculty Fellow, and Director | 2050 Torgersen Hall | 620 Drillfield Drive | VT | Blacksburg VA 24061 | 540-231-1192 | wfeng(a)vt.edu

1 0

Fwd: Data Skills for All Career Accelerator 2024 Program Launch
by Mohammed Seyam 20 Feb '24

20 Feb '24

---------- Forwarded message --------- The DS4A Career Accelerator is a 100% free career advancement program designed for data professionals who are looking to accelerate their careers. We invite 100+ Fellows from the nation’s top universities and research institutes and provide them with training, resume refinement, skill communication, portfolio development with coaches, and mentorship. Additionally, our program matches interested Fellows with potential job opportunities at Fortune 500 companies like Target, Point72, Johnson & Johnson, and Match Group, and connects them to a powerful network of peers and mentors in the data industry. *The program itself is entirely free*, including the application, and by invite-only. Ultimately, our goal is to help our Fellows launch their careers with a powerful industry support structure and develop into the next generation of data-driven leaders. *Data Skills for All Career Accelerator 2024* *When*: March 30th to April 2024 (full schedule *here* <https://www.correlation-one.com/data-skills-for-all?utm_source=amb&utm_medi…> ) *Where*: Virtual *Who*: Working professionals or Master’s/PhD students in transition or looking for new opportunities, with the following *minimum qualifications*: - 2-3 years working in a data role (data analyst, BI analyst, Data Scientist, etc.) OR working with data in their current role (Marketing Analyst, etc.) - Proficiency in one of the following languages: Python, R, SQL, SageMath (must) - Proficiency in one of the following visualization tools: PowerBI, Domo (good to have) - Legally reside and authorized to work in the United States *Cost*: Free, participants are selected based on merit *Apply to DS4A Career Accelerator 2024* <https://www.correlation-one.com/data-skills-for-all?utm_source=amb&utm_medi…> The application will be open until March 18th, but *the deadline for Round 1 applications is February 26th*. We highly recommend you apply as early as you can because our Admissions Committee works on a rolling basis. If you have any questions regarding the program or application process, please send an email to our team at *info.ds4a(a)correlation-one.com* <info.ds4a(a)correlation-one.com>. Best, The Correlation One Team -- Beatrice Khoo Business Operations Manager Correlation One <https://www.correlation-one.com/> 1330 Avenue of the Americas 12th Floor New York, NY 10019 <http://www.correlation-one.com/> -- *Mohammed Seyam <https://website.cs.vt.edu/people/faculty/mohammed-seyam.html>, *PhD Collegiate Assistant Professor Coordinator for Experiential Learning <https://xl.vt.edu/> Department of Computer Science <http://cs.vt.edu/> Virginia Polytechnic Institute and State University (VirginiaTech <http://www.vt.edu/>). Website <https://cs.vt.edu/people/faculty/mohammed-seyam.html>|LinkedIn <https://www.linkedin.com/in/msseyam/> | Facebook <https://www.facebook.com/MSSeyam> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campai…> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campai…> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

1 0