DiscoverMy Favorite TheoremEpisode 85 - Matthew Kahle
Episode 85 - Matthew Kahle

Episode 85 - Matthew Kahle

Update: 2023-06-02
Share

Description

<input onclick="this.value=this.value=='Episode Transcript (click to expand)'?'Hide Transcript':'Episode Transcript (click to expand)';" type="button" class="spoilerbutton" value="Episode Transcript (click to expand)">




Kevin Knudson: Welcome to My Favorite Theorem, the math podcast with no quiz at the end. I'm Kevin Knudson, professor of mathematics at the University of Florida, and I am joined today as always by my fabulous co-host.



Evelyn Lamb: Well, thank you. I'm Evelyn Lamb, a freelance math and science writer in Salt Lake City. And anyone who's on this Zoom, which is only us and our guest, can see that I am bragging with my Zoom background right now. We just got back from a trip to southern Utah, and I took possibly the best picture I've ever taken in my life. And 95% of the credit goes to the clouds because they just — above these red rock hoodoos outside of Bryce Canyon, I turned around and looked at it while we were hiking, and I was like, Oh, my gosh, I have to capture this.




KK: It is quite the picture.




EL: My little iPhone managed.




KK: Yeah. Well, they're pretty good now. Yeah. Anyway, so yeah, I'm getting ready to — I have three trips in the next three weeks. So lots and lots of travel, and I'm gonna make sure I mask up and hopefully I don't come home with COVID, but we'll see.




EL: Yes.




KK: Anyway. So today, we are pleased to welcome Matthew Kahle. Matt, why don’t you introduce yourself?




Matthew Kahle: Hi, everyone. Thanks for having me, Kevin and Evelyn. I'm a mathematician here at The Ohio State University in Columbus, Ohio. I've been here for 11 or 12 years now, and before that, I spent a good part of my life in the western United States. So those clouds look familiar to me, Evelyn. I miss the Colorado sky sometimes.




EL: Yeah, just amazing here.




KK: You did your degree in Seattle, right?




MK: I did. I did my PhD at the University of Washington.




KK: Yep. Yep.




EL: Great. And what is your general research field?




MK: I work a little bit between fields. My main interests are topology, combinatorics, and also probability and statistical physics. And I think I usually feel most comfortable, or maybe I should say most excited mathematically, when there's sort of more than one thing going on, or when it's in the intersection of more than one field.




KK: Yeah, lots of randomness in your work. He’s got this very cool stuff with random topology. And I remember, some paper you had few years ago, I remember really sort of blew my mind, where you had some, you're just computing homology of these random simplicial complexes, and, like, some four- or five-complex had torsion of order, you know, I don't know, 10 to the 12th, or some crazy torsion coefficient. Yeah.




MK: Yeah. So we were really surprised by this too, and we still don't really have any way to prove it, or really understand it very deeply. Kevin was mentioning some experimental work I did with some collaborators a few years ago. But yeah, that is the gist of a lot of what I think about, is random topology, which I sometimes try to sum up as the study of random shapes. And one of the original motivations for this was as sort of a null hypothesis for topological data analysis, that if you want to do statistical methods — if you want to use topological and geometric methods, and statistics and data science, you need a probabilistic foundation. But one of the things we've discovered over the last 15 years or so is that these random shapes are interesting for their own sake as well. And sometimes they have very interesting, even bizarre, properties, where we don't even know how to construct shapes that have these properties at all, but they're they are there. And we know they exist, because of the probabilistic method. Yeah.




EL: So let me be the very naive person who asks, like, how do you, I guess, come up with — like, what do you randomize about shapes? Or you know, if I think about, I don't know, randomly drawing from from some sort of, I don't know, bucket of properties, is it that or is it… Just what is random? What quantity or quality is being randomized?




MK: Right. So a lot of the random shapes or spaces that I've studied have have been on the combinatorial or discrete side. So for example, there are lots of different types of random simplicial complex that people have studied by now. And typically, you have just some probability distribution, some way of making a random simplicial complex on n vertices. And n can be anything, but then the yoga of the subject is that typically n goes to infinity. And then we're interested in sort of the asymptotic properties as your random shape grows. So one of the early motivations, or early inspirations, for the subject of random, simplicial complexes was random graph theory. So you can create random networks various ways, and people have been studying that for for a bit longer, probably at least 60 years or so now, with new models and new interesting ideas coming along all the time. For example, there was originally the Erdős–Rényi model of random graph where the edges all have equal probability, and they're all independent. This is a beautiful model mathematically, and it's been studied extensively. We really know lots and lots about that model of random graph now, although surprisingly, people can continue to discover new things about it as well. But in today's world, some people have studied other models of random graphs that they say may have made better model real world networks, for example, social networks, or what we see in epidemiology, and so on. The Erdős–Rényi model is something that's tractable, and that we can prove deep math theorems about, but it might not be the best model for real world networks. But, you know, I think of the random simplicial complexes that I study sometimes as just higher-dimensional versions of random graphs.




EL: Okay.




MK: So as well as as well as vertices and edges, we can have higher-dimensional cells in there, and and that starts to sort of enrich the space. It's not just one-dimensional now, it could be two-dimensional, or it could be any dimension.




EL: So you might not know. You've got some large number n, and you might not know what dimension this random — you’re, like ,attaching with edges with some sort of probability between any two things. And so you might not know what dimension your simplicial complex is going to be until after you randomly assign all of these edges and faces and, you know, and n- whatever the word is for that, n-things. [Editor’s note: it’s n-simplex.]




MK: Yeah, absolutely. That's right. It could be that the dimension of the random simplicial complex is itself a random variable. And you know, that we don't ahead of time even know what the dimension of it is.




EL: Cool!




KK: So there's lots to do here. This is why Matt has lots of students and lots of lots of good projects to work on. But anyway, we invited you on not just to talk about this really interesting mathematics, but to find out what your favorite theorem is. So what is it?




MK: Okay, so I've been thinking about this. Well, I have to admit, I think I asked myself this just knowing of your podcast in case I ever got invited on. And then I've been thinking about it since you invited me. I would say my favorite math theorem, probably the one I've thought about the most, the one maybe that affects me the most, is Euler’s polyhedral formula, which is V−E+F=2. Right? So let's just start out saying, well, you know, what do we mean by this? I think my understanding of the history of it is that it was something that as far as we know, the Greeks didn't observe even though they were interested in convex polyhedra. And sometimes people consider the classification of the perfect Platonic solids is one of the peak contributions of Euclid’s Elements. But we don't know that they recognized this pattern that Euler noticed thousands of years later. If you take any convex polyhedron, a cube or an icosahedron, or a pyramid, a bi-pyramid, any kind of three dimensional polyhedral shape that you can imagine that's convex, V, the vertices is the sort of number of corners of the shape and E is the number of edges. And then F is the faces. It always is the case that V−E+F=2. So Euler noticed this. And it's not clear if he gave a rigorous proof or not. I don't even know if he felt like anything needed to be proved, maybe it was obvious to him. And nowadays, we have many, many beautiful proofs of this fact. But one of the things that strikes me about it is that, it’s sort of in hindsight, is that this is just sort of the tip of a very big iceberg. There's a much more general fact that we are just kind of getting our first glimpses of, and nowadays, we would think of this as not just a phenomenon about convex polyhedron, 3-dimensional space, that it’s just a general phenomenon in algebraic topology, or you can say =more generally, in homological algebra. It's just sort of a feature of nature somehow.




KK: Right, right.




EL: I think something that I really enjoy about this fact is you can present it at first as a theorem or as a fact. But then this fact kind of leads you to this new definition that you can observe about all sorts of different shapes, you know, this number that is the vertices minus the edges plus the faces, hopefully, I got it in the right order, yes. Then you can assign that, you know, you can say, like, what does, you know, if you've got a torus, like a polyhedral torus, or, you know, a higher-genus object or a higher-dimensional thing, you can sort of use this, and so it's like a fact becomes a definition or a new thing to observe.




MK: That’s right. Are you saying, for example, you know, we have the
Comments 
00:00
00:00
x

0.5x

0.8x

1.0x

1.25x

1.5x

2.0x

3.0x

Sleep Timer

Off

End of Episode

5 Minutes

10 Minutes

15 Minutes

30 Minutes

45 Minutes

60 Minutes

120 Minutes

Episode 85 - Matthew Kahle

Episode 85 - Matthew Kahle

Kevin Knudson & Evelyn Lamb