TechBytes (formerly CS/DS Tea) is a weekly department event where students and faculty gather to discuss computing related topics. Tea, cookies, and (sometimes) pizza are served. Below is a list of the special events planned for the Fall 2023 semester! New attendees are always welcome!
Unless otherwise noted, TechBytes events usually take place on Thursdays at 11:30am-12:30pm in Ford 102.
Interested in presenting at TechBytes? Please contact Haiyan Cheng at email@example.com. To sign up for news about TechBytes events, or for other questions, contact Lizzi at firstname.lastname@example.org.
Pete Wirfs, "My Career in Information Technology"
In the 1970s, educational paths were linear, constraining students to predetermined careers. The 1980s introduced adaptive learning through on-the-job training due to scarce software solutions. The 1990s celebrated the Mainframe's supremacy, while Y2K concerns cast doubts. The 2000s witnessed the decline of the Mainframe era. In the 2010s, Mainframes powered down, yielding to Oracle and .NET dominance. Now in the 2020s, focus turns to both the future and the fate of Mainframes.
Pete worked as a software engineer for SAIF Corporation and recently retired after 42 years of service. He grew up in Corvallis and in 1979 completed his OSU degree in Business Administration with a minor in Computer Science. His son is currently a student at
Bernadette Boscoe, "Arrays from the Sky: Astronomy and Data Science/Computer Science"
Machine learning is increasingly used in astronomy to process the massive amounts of data collected from new, powerful telescopes. Computer scientists work with astronomers to build software and pipelines to enable scientific research. In this beginner-level talk, Dr. Boscoe will explain how machine learning and machine learning pipelines work to help us understand the evolution of our Universe. Her research looks at opportunities and challenges in building computational systems, resulting in best practices for astronomers and infrastructure builders.
Dr. Boscoe is an Assistant Professor of Computer Science at Southern Oregon University. She did her graduate work at UCLA, and holds an MS in Mathematics from California State University, Northridge. Prior to this, she worked as a technologist at Cornell University and back in the day was a full stack engineer.
Student Internship Talks
Hriday Raj, "Machine Learning Internship Experience"
Rachel Brown, "What You See Where You're Not Looking", Ford 101
Although we have a rich perception of the world around us, human vision is not uniform everywhere. Our peripheral vision provides us enough information to understand the gist of a scene, but when we probe the details of that perception we find them lacking. In this talk, Dr. Brown will describe a phenomenon called “crowding”, which limits our ability to perceive certain details in our peripheral vision. Her exploration of these topics will focus on how we can use machine learning techniques to build models of human perception and action that are practical for both vision science and computer graphics research.
Dr. Rachel Brown is a researcher studying the intersection of human perception, computer graphics, and machine learning. She earned her B.S. in Biology and Psychology at The College of William & Mary in Virginia and later went on to complete an M.S. in Computer Science and a Ph.D in Vision Science from the University of California at Berkeley. Dr. Brown spent over six years working in industry as a Senior Research Scientist at NVIDIA before transitioning back to academia. She is currently teaching Computer Vision for the UC Berkeley MIDS program, as well as Data Science and Machine Learning at Lewis & Clark College in Portland, Oregon.
Sam Schwartz, "Machine Learning Applications"
Machine learning is an increasingly important part of our research ecosystem. In this talk, Sam will discuss two key areas of his current scholarship. First, Sam collaborates with various scientists to create enabling tools and technologies, usually driven by complex machine learning and graph-theoretical algorithms. He also engages in statistical analysis of (often messy) domain-specific datasets. Second, Sam’s PhD dissertation work involves empirical quantitative analysis of research software engineering projects, particularly within the advanced scientific computing space. He aims to answer questions such as, “Just how many scientific software projects are out there? What do they look like? How do they compare to other types of software?”
Sam Schwartz is a Ph.D. candidate in the Department of Computer Science at the University of Oregon. Sam earned his master's degree in Mathematics from Utah State University, where he also earned his bachelor's degree as a double major in Mathematics and Computer Science with minors in Spanish and Organizational Communication. Sam has also spent time working with a variety of federal government national laboratories, small businesses and large technology companies, and teaching high school English in Chile. When not working, Sam enjoys driving up and down Oregon's beautiful coast with his dog.
Jed Rembold, "Harnessing the Hidden Potential of Databases"
In academia and the scientific realm, we constantly grapple with the need to store and retrieve vast amounts of information. Often, the go-to solution for managing this data is the familiar spreadsheet. However, more complex data can often resist the confines of a single tabular structure, leading to data manipulation headaches and integrity issues. In this talk I will champion the widespread adoption of databases for a variety of professional and personal projects. We'll explore the distinct advantages, potential pitfalls, and offer practical use-cases and examples. Some projects will still require a spreadsheet interface to the data, and so we'll also look at ways to bridge the gap between spreadsheets and databases. Our goal is to leave you inspired and equipped with the knowledge and tools to begin harnessing the power of databases in your own projects. Embrace a more robust, flexible, and efficient approach to managing your information!
Jed Rembold is a professor in the computer science department at Willamette University.
Mike Woodford, "Data Modernization: Challenges and Opportunities in the Public Sector"
Government agencies have traditionally held the roles of preeminent suppliers, and sophisticated ethical users, of data. Today’s challenge is for government entities to proactively maintain these roles, or they will struggle to fulfill their civic duties and carry out their missions. Governments have, over decades, acquired massive volumes of data assets through the investment of hundreds of millions of dollars and hundreds of thousands of people-hours …and in the process incurred increasing amounts of data debt. We are at an inflection point. Data are business assets, and data modernization is a business challenge. New roles and skills are necessary to overcome the massive data debt burden and to operationalize vast data stores in new ways that have significant benefit to agencies and organizations.
Mike Woodford is the Chief Data Officer at the Oregon Department of Transportation. See his LinkedIn profile.
MONDAY, November 20th
Deahan Yu, "Understand and Improve Health from Text", Ford 101
In the era of big data, the richness of text offers great potential for better understanding and advancing health. Natural language processing (NLP), including deep learning, offers the potential to extract insights from health-related text data. This talk will demonstrate the application of NLP to extract clinical features from unstructured text data in two different contexts: clinical notes and user-generated text. This talk will also outline the potential impact of uncovering health information from text data.
Deahan is currently a PhD candidate in the School of Information at the University of Michigan. Overall, his research aims to better understand health and advance healthcare through data-driven approaches. Deahan's work utilizes natural language processing and machine learning to extract clinical features from health-related text data.
MONDAY, November 27th
Mian Adnan, "Time Series Prediction: Machine Learning versus Deep Learning", Ford 101
The performance of the existing neural network models for time series predictions is evaluated
Mian Adnan is a PhD candidate at Bowling Green State University. He has masters degrees in actuarial science from Ball State University and statistics from Indiana University. His research is based on machine learning, deep learning, data science, high dimensional statistics, and financial engineering. Mian has over 60 published papers from various journals and conference proceedings and has taught several courses in statistics and bioinformatics at both the undergraduate and graduate levels. Outside of the classroom, Mian enjoys singing and performing.
Graduate Programs Info Session
Led by Hillary Patterson (Director of Recruiting and Corporate Outreach), this session will cover information about the Masters in Data Science and the Masters in Computer Science. All interested students are invited to attend.