CCR AI/ML Seminar: Machine Learning Approaches for Characterizing Global Sea Surface Temperature Fields

The Center for Computational Research is planning a series of monthly talks on AI/ML techniques and applications across various scientific domains. We envision the talks to strike a good balance between depth and breath. The goal of these talks will be to (i) introduce the particular AI/ML technique to fellow faculty and graduate students who have a basic understanding of deep learning and (ii) present a variety of applications in different domains without assuming deep domain knowledge. Each event will start with a 45 minute talk with 15 minutes for questions and a subsequent 30 minute networking session for brainstorming and further discussion. The speakers will be URI faculty from a number of colleges.

Details on the first talk appear below.

Speaker: Peter Cornillon (GSO)
Date/Time/Location: May 1st, 3pm, 112 East Hall.
Title: Machine Learning Approaches for Characterizing Global Sea Surface Temperature Fields
Abstract: Sea surface temperature (SST) fields derived from satellite-borne sensors offer an ideal dataset for exploration using machine learning techniques. In this presentation, I will describe the use of an auto-encoder, in combination with a flow equalization technique, to identify outliers in a 20-year, global, twice-daily archive of SST fields. I will then demonstrate how this same approach can be applied to evaluate the performance of a global ocean circulation model.
Switching gears, I will introduce a machine learning model based on contrastive learning applied to the same SST datasets—this time to uncover and categorize fundamental spatial patterns within the fields. Finally, I will briefly touch on an analysis of the latent space produced by the contrastive learning model, with a focus on estimating the intrinsic dimensionality of SST field variability.

 

AI Big Data Forum — Spring ’25

April 18th noon, Galanti Lounge. Details in the attached flyer! 

Event Flyer

URI Speed Networking Event for Graduate & Senior Students

Invitation: URI Speed Networking Event for Graduate & Senior Students (AI, Quantum, and Computationally Enhanced Research & Innovation) 

Friday April 11, 2025, 9:30-noon

Galanti Lounge, 3rd Floor, URI Library

by

Center for Computational Research, ITS (Joan Peckham & Gaurav Khanna)

URI Innovation Lab (Jim McGwin)

URI Division of Research and Economic Development (Karen Markin)

PURPOSE: To assist graduate and senior students engaged in research and/or innovation at URI to find collaborators from across campus, and to attend to increased need for interdisciplinary, convergence1, use-inspired2, and data enabled research and innovation.

APPROACH: This event will provide a means for students of all disciplines who are active in research and/or innovation to meet and find collaborators from other disciplines. This includes:

  • Student researchers or innovators of any discipline searching for partners with analytical, and/or computational expertise, or
  • Student researchers or innovators with applied and/or theoretical data, quantum computing, mathematics, statistics or AI/analysis/management expertise wishing to partner with researchers and/or innovators from other disciplines.

The goal is to explore interdisciplinary research collaborations that can strengthen research/innovation projects and better prepare participants for careers that increasingly call for interdisciplinary data and computationally enhanced collaborations.  URI students of all disciplines who are actively engaged in research and innovation are invited.

PROCESS: We will use a structured and timed speed dating approach that will give each scholar short segments of time to exchange their expertise, interests, and innovative ideas with other participants. At the end we will provide a pizza and beverages, but all will be welcome to stay and further discuss promising collaborations.

Registration by Friday, April 4, 2025 (So that we can order food): https://docs.google.com/forms/d/e/1FAIpQLSflirdAKzDdSL-iJiU9m38gcQ1W3dTua78YkJTt9aB78AdnlQ/viewform?usp=dialog

Questions? Contact jpeckham@uri.edu (Joan Peckham)

 

[1] Convergence Research (https://beta.nsf.gov/funding/learn/research-types/learn-about-convergence-research): It is driven by a specific and compelling problem, whether that problem arises from deep scientific questions or pressing societal needs. It shows deep integration across disciplines. Convergence research intentionally brings together intellectually diverse researchers to develop effective ways of communicating across disciplines. As experts from different disciplines pursue a common research challenge, their knowledge, theories, methods, data and research communities increasingly intermingle. 

[2] Use-Inspired Research – From NSF solicitation on AI Institutes – NSF 22502 (https://www.nsf.gov/pubs/2022/nsf22502/nsf22502.htm):  We use the phrase “use-inspired” rather than “applied” to emphasize that this solicitation seeks to support work that goes beyond merely applying known techniques and adds new knowledge and understanding in both foundational AI and use-inspired domains. Ideally there is a virtuous cycle between foundational and use-inspired research, where foundational results provide a starting point for use-inspired research, and the results from use-inspired research are generalized and made foundational.

Spring Semester Workshops by URI CCR / AI Lab

Welcome back! The details of our CCR / AI Lab semester workshops are presented below. These are open to faculty, staff and students and are totally free. 

https://docs.unity.uri.edu/news/2025/01/uri-spring-25-workshops/
You can also find these on the URI events calendar (and subscribe): https://events.uri.edu/group/ai

We have some relatively new workshop offerings on AI Tools, Gen AI and Bioinformatics. A huge thanks to Indrani Mandal and our student team for planning and preparing these workshops!

Major Unity Upgrade!

We are pleased to announce that a major upgrade to our primary HPC/AI platform UNITY is finally complete! We have added:

1,000 CPU-cores: 16, 64-core CPU nodes that are identical to our current nodes in the uri-cpu partition.
24 AI GPUs: 4 nodes with 4 Nvidia L40S GPUs each; 2 nodes with 4 Nvidia H100 GPU nodes.

Moreover, snapshots are available for /project, in addition to /work and /home; so you can do file recovery. And there is 500TB more scratch space available at /scratch3. Finally, new documentation has been added with a new “Get Help” section.

Please don’t hesitate to reach out for assistance with these!

In today’s classrooms, the availability of Artificial Intelligence (AI) is impacting learning—it is a time of raising questions as AI provides a variety of options and insights. AI tools can now assist students in writing, problem-solving, and research; offering new levels of convenience and access. But, while these tools can be helpful, they also bring challenges. As educators, it’s crucial to not just introduce AI, but to cultivate an environment of critical thinking that balances both caution and curiosity, all the while empowering students to ask: What might be missing? Can I trust this output? and What does it mean to use AI responsibly?

Critical thinking has long been a cornerstone of higher education and intellectual development. In an AI-enabled world, students need to learn to go beyond using these tools; they need to understand how to generate insights through verifying information, as well as understanding the possibilities and limits of technology.

A group of students viewing screen togetheter vith AI background theme

When students engage in critical thinking with AI, they’re better prepared to:

  • Think creatively and independently: Critical thinking encourages students to consider multiple perspectives and solutions, rather than simply relying on AI-generated answers. This independence nurtures innovation and personal insight.
  • Distinguish fact from fabrication: While AI can generate vast amounts of text, not everything it produces is accurate. Encouraging students to fact-check and cross-reference helps cultivate a healthy skepticism.
  • Challenge assumptions: AI often reflects only its training data. By guiding students to analyze the sources (including question potential biases) and recognize how assumptions shape information, it can help foster critical thinking.

Key areas to explore in AI’s limitations include:

  • Accuracy and Misinformation: AI produces results based on patterns in data rather than true understanding. Students may mistake plausible-sounding, yet incorrect information, for fact, undermining their knowledge and learning integrity.
  • Data-Driven Biases: AI systems inherit biases from the data used to train them, potentially perpetuating skewed perspectives. Encouraging students to question these biases nurtures an awareness of how assumptions shape content, fostering a more discerning, balanced view of information.
  • Risks to Independent Thought: Over-relying on AI can hinder a student’s own critical thinking skills. While AI might offer shortcuts, true learning often comes from grappling with complexity, not from accepting easy answers.

Ultimately, while AI may seem to provide quick solutions, it cannot replace critical thinking. Many AI-generated responses appear confident and well-formatted, however although outputs may: miss nuance, need detailed fact checking, or reflect underlying biases from source materials. Approaching AI materials with critical thinking can help students in recognizing these pitfalls and develop habits of inquiry that can prevent them from adopting AI’s suggestions without expert review.

As an institution of higher education, we have the ability to foster a mindset of inquiry. Consider the following strategies to help students and ourselves engage thoughtfully and critically with AI:

  1. Encourage Source Verification: Just as we ask students to cite sources in their own work, we can guide them to question AI sources and verify AI-generated content. This practice reinforces the importance of credible information and builds a habit of checking facts.
  2. Examine AI’s Limitations Together: Bring AI-generated outputs into class discussions; exploring where they succeed and where they fall short. This exercise helps students recognize that AI’s “knowledge” is limited, often lacking the context, depth, and human judgment necessary for complex analysis.
  3. Practice “Spot the Error” Activities: Regularly review AI outputs in class to identify inaccuracies, ethical concerns, or biases. This approach not only develops a critical eye, but reinforces the idea that AI should be questioned and evaluated, and not just passively accepted.
  4. Engage in Ethical Dialogues: The ethics of AI use extend beyond academic integrity, it also includes privacy issues and potential societal impacts. Encouraging students to reflect on these implications fosters a responsible mindset, helping them consider the broader impact of their technology use.
Critical Thinking and Ai image

 

While AI tools can provide new educational possibilities, there’s value in asking when to use AI. Asking the question of whether AI truly serves the learning objectives of the course and assignment. Some lessons may be better learned by working through challenges without automated assistance, promoting creativity, resilience, and deep, independent analysis. However, by selectively incorporating AI, educators can help students appreciate it as a tool that, while powerful, doesn’t replace the need for human insight and critical judgment.

Encouraging thoughtful reflection with a little skeptism toward AI helps students maintain their intellectual independence. Rather than seeing AI as a replacement for their own reasoning, they’ll learn to use it as a complement to their critical thinking. This balanced approach supports a learning environment where technology is seen as a helpful aid but not an unquestionable authority.

In an AI-enhanced world, it’s more important than ever to cultivate critical thinking and intellectual independence in students. Through a balanced approach—one that blends curiosity with caution —we can empower students to use AI responsibly and thoughtfully. Let’s encourage our students to ask questions, challenge outputs, and think critically so that, no matter where technology advances, they are equipped with critical thinking, curiosity, and insight.

 

AI Big Data Forum: Quantum Computing and AI

Two speakers:

Len Kahn, Chair of Physics, URI – Hear how we are preparing for quantum computing at URI

Stephen Bach, Computer Science, Brown University:

It’s All About Data: The Promises and Limitations of Recent Developments in AI 

Talk Abstract: This talk will overview the evolution of AI over the last five years, through the lens of machine learning and large language models. Accessible to scientists with a general computing background, we will discuss the key technical developments that have enabled recent advances. Many of them are data-centric, meaning that the development of datasets has been at least as important as advances in model architectures and algorithms. The centrality of data means that further AI advances are also limited by data, particularly in specialized domains requiring subject matter expertise. I will discuss these challenges and share some of our lab’s recent work on overcoming them.

Bio: Stephen Bach is an assistant professor of computer science at Brown University. His research focuses on weakly supervised, zero-shot, and few-shot machine learning. The goal of his work is to create methods and systems that drive down the labor cost of AI. He was a core contributor to the Snorkel framework, which was recognized with a Best of VLDB 2018 award. Snorkel is used in production at numerous Fortune 500 companies for programmatic training data curation. He also co-led the team that developed the T0 family of large language models. The team was also one of the proposers of instruction tuning, which is the process of fine-tuning language models with supervised training to follow instructions. Instruction tuning is now a standard part of training large language models. Stephen is also an advisor to Snorkel AI, a company that provides software and services for data-centric AI.

Nov 22nd 12pm Galanti Lounge (Library)

 

Speed Networking Event AI, Quantum, and Computationally Enhanced Research & Innovation

Invitation, URI Speed Networking Event

AI, Quantum, and Computationally Enhanced Research & Innovation  

Friday October 11, 2024, 9:30-noon

Galanti Lounge, 3rd Floor, URI Library

Center for Computational Research, ITS (Joan Peckham & Gaurav Khanna)

URI Innovation Lab (Jim McGwin)

College of Business, Business Analytics and AI (Drew Zhang)

 URI Division of Research and Economic Development (Karen Markin)

 

PURPOSE: To support scholars at URI to respond to solicitations that increasingly request interdisciplinary, convergence1, use-inspired2, and data enabled research. 

APPROACH: This event will provide a means for scholars of all disciplines with computational needs, and applied & theoretical data, quantum computing, mathematics, statistics and AI/analysis/management scholars to meet and explore interdisciplinary research collaborations that can result in publications, external funding and/or startups.  URI scholars and students of all disciplines who are actively engaged in research and innovation are invited.

PROCESS: We will use a structured and timed speed dating approach that will give each scholar short segments of time to exchange their expertise, interests, and innovative ideas with other participants. Snack and beverages will be available upon arrival. At the end we will provide a grab and go lunch, but all will be welcome to stay and further discuss promising collaborations. 

Registration by Friday, October 4, 2024 (So that we can order food): https://forms.gle/YHsFW3bRhhjgsw7B8

Questions? Contact jpeckham@uri.edu (Joan Peckham)

 

1 Convergence Research (https://beta.nsf.gov/funding/learn/research-types/learn-about-convergence-research): It is driven by a specific and compelling problem, whether that problem arises from deep scientific questions or pressing societal needs. It shows deep integration across disciplines. Convergence research intentionally brings together intellectually diverse researchers to develop effective ways of communicating across disciplines. As experts from different disciplines pursue a common research challenge, their knowledge, theories, methods, data and research communities increasingly intermingle. 

 

2 Use-Inspired Research – From NSF solicitation on AI Institutes – NSF 22502 (https://www.nsf.gov/pubs/2022/nsf22502/nsf22502.htm):  We use the phrase “use-inspired” rather than “applied” to emphasize that this solicitation seeks to support work that goes beyond merely applying known techniques and adds new knowledge and understanding in both foundational AI and use-inspired domains. Ideally there is a virtuous cycle between foundational and use-inspired research, where foundational results provide a starting point for use-inspired research, and the results from use-inspired research are generalized and made foundational.

 

—