# Past Academic Seminars - 2022/23

## Block 8 Seminars

#### Friday May 5

**12:30 - 1:30 PM, Kresge Lecture Hall (Tutt Science 122) Dr. Iris Bahar - “Energy-efficient, Reliable (& Fun!) Computing Across the Hardware/Software Stack”**

Abstract:

Prof. Iris Bahar has been working on the design of computer systems for the past 3 decades. Her research focuses on developing new approaches to reduce power dissipation and improve reliability in high-performance processors, specialized embedded systems, and computing systems designed with emerging technologies. Her recent interests have led her to consider design of machine learning and robotic systems, and how they can benefit from energy-efficient design techniques. In this talk, Prof. Bahar will give an overview of some of her current research projects covering memory design techniques using near-memory-processing, and efficient use of machine learning techniques for robust scene perception. She will also talk about her creation of new interdisciplinary courses that mix art, design, engineering, and computing. Finally, she will conclude with her thoughts on graduate school. She is happy to make this an interactive talk so feel free to come ready to ask questions.

Rated PG: Some undergraduate computer science recommended.

#### Tuesday May 9

**12:30 - 1:30 PM, Kresge Lecture Hall (Tutt Science 122) Dr. Kurt Fleischer - “Tales of Computer Science Research in the Animation Industry”**

Abstract:

This talk will give a personal view of research projects done in service of computer animated movies. It will cover a few projects that have been published at Siggraph describing the technology behind methods for stylized animation, per-shot sculpting of silhouettes, and a new rigging method based on profile curves.

Rated PG: Some undergraduate computer science recommended.

#### Friday May 12

**12:30 - 1:30 PM, Kresge Lecture Hall (Tutt Science 122) Dr. Jennifer Marlow - “How to Automate Without Losing Your Humanity: Insights from research on designing human-centered automation”**

Abstract:

Automated tools and features on mobile phones, computers and other devices can help people perform tasks quickly and efficiently, from making phone calls on our behalf to suggesting automatic chat replies. But should we automate a task just because we can? This talk will cover a general framework for thinking about human-centered automation: What factors influence the perceived value of automation, and how can the design of automated systems can help build trust when things don't always go as expected? We'll then explore specific use cases related to automatic app management to see how people feel about features providing assistance in this space and how we might make these features even more useful.

Rated PG: Some undergraduate computer science experience may be helpful, but the talk should be fairly accessible!

#### Wednesday May 17

**3:00 - 4:00 PM, Kresge Lecture Hall (Tutt Science 122) Dr. Robin Wilson - “The Origins of Mathematics”**

Abstract:

Where did mathematics come from? What number systems were used? What calculations were carried out, and how accurate were they? What problems were tackled, and why? What forms did arithmetic, geometry, and algebra take in ancient times? This talk introduces the varying mathematical activities of ancient Egypt and Mesopotamia before proceeding to the mathematics of ancient China and India and the calendar concerns of Mayans of central America.

Rated PG: Some undergraduate math experience may be helpful, but the talk should be fairly accessible!

#### Friday May 19

**1:30 - 2:30 PM, Tutt Science 213 Dr. Robin Wilson - “Graph Theory in America: The First Hundred Years”**

Abstract:

Graph theory is the mathematical study of connections, as may be encountered, for example, in road, airline, or telecommunications networks. This talk covers the development of the subject in America’s second century from 1876, when James Joseph Sylvester arrived in the US and introduced the word “graph” (in this sense) in connection with his work on mathematical trees and chemical molecules, to 1976, when the longstanding four color problem on the coloring of maps was finally solved. In particular, this talk will introduce the personalities involved, as described in the recently published book “Graph Theory in America” (Princeton University Press, January 2023), two of whose authors are myself and CC’s Emeritus Professor John Watkins. No particular knowledge of graph theory is assumed.

Rated G: No particular knowledge of math needed!

## Block 7 Seminars

#### Friday April 7

**3:00-4:00pm, Tutt Science 229 Prof. Cory Scott - “Towards Differentiable Graph Comparison”**

Abstract:

Graphs are a natural way to capture data about structure - that is, how a bunch of objects are related to one another. But what if we want to compare one graph to another? For example, detecting whether a social network is comprised of real humans versus chatbots. Testing two networks for similarity turns out to be a really difficult problem in theoretical computer science - but there are potential applications to a wide variety of applications including biology and medicine. In this talk, I'll discuss what makes this problem difficult. I'll also talk about ways to apply machine learning to the problem, as well as what major roadblocks remain in machine-learning enabled graph similarity scoring.

Rated PG-13: Some calculus and linear algebra required to understand everything.

#### Friday April 14

**12:30-1:30pm, Kresge Lecture Hall (Tutt Science 122) Prof. Stefan Erickson - “Game of Codes”**

Abstract:

A chill wind was blowing from the north. Somewhere in the distance, a wolf howled. Eddard Stark shivered. "Winter is coming," the Lord of Winterfell said. "I must tell the king."

Maester Luwin considered his lord's words. "We must be cautious. Lord Varys has eyes and ears everywhere. We cannot allow this knowledge to fall into the hands of the Lannisters."

"What do you have in mind?" Lord Stark asked his most trusted advisor.

"We must create a secret language that only you and King Robert can understand," the maester replied.

"But how? Robert is many leagues away in King's Landing. We could send a raven, but ravens can be intercepted. Even our most trusted messengers can be captured and tortured."

Maester Luwin smiled. "Never fear, my lord. There are new advances in the craft of enciphering that cannot be broken with all the gold in Casterly Rock. Allow me show you how."

Rated PG: Some familiarity with modular arithmetic and George R. R. Martin's "A Song of Ice and Fire" is recommended.

## Block 6 Seminars

#### Friday Mar 3

**3:00-4:00pm, Tutt Science 229 Prof. Joseph Rennie - “What is a Galois Theory? Math as a Structure Through a Homotopical Lens”**

Abstract:

In this talk about math, structures invented to analyze topological spaces emerge as the proper lens for viewing Galois Theory. The goal of this talk will be to give a sense of what it is like to know homotopy theory and higher category without requiring the audience to know either. We will see how naturally curiosity can take a real-world interest to a deep and gratifying insight about the structure of math itself. Of course, we will conceptually cover Galois Theorems multiple times over as we attempt to unify them under one result.

Rated PG-13: Considerable undergraduate math experience recommended.

#### Friday Mar 10

**12:30-1:30pm, Kresge Lecture Hall (Tutt Science 122) Prof. Cory Scott - “Computer Scientists Need Derivatives Too: An Introduction to Optimization”**

Abstract:

Across all fields of science and engineering, we frequently need to optimize a function: we have some function f, and we want to find a choice of x that will make f(x) as large as possible. f can be a mathematical function, or something more complicated, like the error between a machine learning model's prediction and the "right" answer. Optimization is a really broad field with many different techniques for specific kinds of functions. In this talk, I'll introduce some examples of optimization algorithms, and demonstrate how they work. I'll also discuss what makes a particular algorithm suitable for specific kinds of problems. If I have time, at the end I'll talk about a couple of optimization algorithms that crop up in my research.

Rated PG: Some calculus knowledge suggested but not required.

#### Friday Mar 17

**2:30-3:30pm, Tutt Science Center 229 Dr. Steve Getty - “Stay Motivated! Understanding and Monitoring Student Motivation in College”**

Abstract:

Motivation is essential in students’ academic performance, engagement, and persistence, and a body of work has laid a foundation for the role that individual motivation plays in educational contexts. This is especially important when students doubt their ability, see little value for engaging in classroom activities, or feel their efforts for success come at too high of a price. While many motivational theories, models have been proposed in such settings, Expectancy-Value Theory provides a comprehensive approach for understanding student achievement and persistence. In this talk, I will 1) *briefly* share the rationale for 3 factors for student motivation in STEM (CFA, Confirmatory Factor Analysis), 2) share data on measuring student motivation in courses at Colorado College, and then 3) discuss implications for students in understanding, supporting their motivation (….and possibly how instructors can support motivation in their courses).

Rated G: No mathematics or computer science background required!

## Block 5 Seminars

#### Friday Feb 17

**3:00-4:00pm, Tutt Science 221 Prof. Molly Moran - “Cutting Sequences in Euclidean and Hyperbolic Geometry”**

Abstract:

Our talk is motivated by a simple question: how do we calculate the slope of a line? Wait...isn't the answer just as simple as rise over run? Yes! But to calculate rise over run, we need at least two points that we know lie on the line. If we are only given a graph and there are no integer points on the line, finding these two points can be challenging. In this talk, we will introduce cutting sequences of lines and show how they can help us find slope. We will also explore what type of information these cutting sequences provide if we switch to hyperbolic geometry.

## Block 4 Friday Seminars

#### Each Thursday and Friday Dec 1 - 16

This block our talks will be presented by three prospective faculty candidates.

**Each THURSDAY there will be a research focused talk from 3PM to 4PM.**

**Each FRIDAY there will be a student focused talk from 12:30PM to 1:30PM (with pizza at noon!)**

These talks are write-up eligible for our majors, so please watch for more information on the topics and location. See you there!

## Block 3 Fearless Friday Seminars

#### Friday Oct 28 2022

**2:30-3:30pm, Tutt Science 229 Prof. Cory Scott - “Graph Lineages and Skeletal Graph Products”**

Abstract:

Machine learning models are limited by the "curse of dimensionality": we only have a fixed amount of memory and computing power, so we have to decide where to use it effectively. As a specific example, machine learning models for images often have to deal with a tradeoff between spatial resolution (how many pixels are in the image) and semantic resolution (how complex an idea our model can express). Many machine learning models get around this tradeoff by gradually shrinking the spatial resolution, while at the same time increasing the feature resolution. In my talk, I will explain how these types of model growth and shrinkage can be described by "graph lineages": growing sequences of related graphs. I will also define "skeletal" graph products, which are a way to combine graph lineages to get another lineage. I will demonstrate how we can use skeletal graph products to build some well-known machine learning models which have the same spatial vs. semantic tradeoff described above. This talk will touch on graph neural networks, clustering algorithms, linear algebra, and matrix decomposition. The last few minutes of the talk will be devoted to brainstorming potential applications of this model architecture.

#### Friday Nov 4 2022

**12:30 to 1:30PM: location TBADr. Rebecca Mitchell**

This CC alum who will be visiting next block to teach Linear Algebra works at the National Renewable Energy Laboratory (NREL) as a researcher, info here. This talk is **write up eligible** and more specifics to come!

#### Friday Nov 11 2022

**3:00 to 4:00PM: location TBAProf. Flavia Sancier-Barbosa**

Widely admired Colorado College mathematics professor Flavia Sancier-Barbosa has a passion for stochastic processes, probability, applied data analysis, and mathematical finance, but more details about this **write up eligible** talk to come!

## Block 2 Fearless Friday Seminars

## Block 1 Fearless Friday Seminars

#### Friday Sep 2 2022

**12:30pm in Tutt Science Lecture Hall (TSC122)**

Math and CS Faculty Research Interests

Followed by Ice Cream Social outside in 1:30pm

#### Friday Sep 9 2022

**12:30pm in TSC 122 Lecture Hall**

*Pizza at noon outside Tutt Science*

Dr. Justin Cole - Dept. of Mathematics, UCCS: An Introduction to Some Rogue Wave Models

*Rating*

*PG-13. Considerable undergraduate mathematics or computer science assumed*