Skip to the Main Content

Note:These pages make extensive use of the latest XHTML and CSS Standards. They ought to look great in any standards-compliant modern browser. Unfortunately, they will probably look horrible in older browsers, like Netscape 4.x and IE 4.x. Moreover, many posts use MathML, which is, currently only supported in Mozilla. My best suggestion (and you will thank me when surfing an ever-increasing number of sites on the web which have been crafted to use the new standards) is to upgrade to the latest version of your browser. If that's not possible, consider moving to the Standards-compliant and open-source Mozilla browser.

February 21, 2007

Cake Talk

Posted by David Corfield

I’ve been a little quiet in the Café of late as I’ve been preparing a departmental ‘cake’ talk. This excellent institution requires last week’s speaker to provide cake, thereby assuring a good audience. You can access the slides here. It’s probably not overly comprehensible (nor perhaps was it even to most of those attending) as it ambitiously gathers together all my interests with regard to machine learning: Bayesianism, information geometry, and nonparametric statistics. I thoughtfully left out what I’ve been learning about probabilistic monads. Posted at February 21, 2007 3:11 PM UTC

TrackBack URL for this Entry:

5 Comments & 0 Trackbacks

Re: Cake Talk

It may’ve been a cake talk, but it sure wasn’t a cakewalk — either to give, or to follow. Intense stuff!

Posted by: John Baez on February 22, 2007 5:47 AM | Permalink | Reply to this

Re: Cake Talk

I’ve added a little to it, and might write an expository paper at some point. Intriguingly, there’s a connection between what’s going on here and those Ricci flows Urs was telling us about. See page 29 of Carlos Rodriguez’ A Geometric Theory of Ignorance.

Is anything known about flows on manifolds which are locally modelled on reproducing kernel hilbert spaces?

Posted by: David Corfield on February 22, 2007 4:17 PM | Permalink | Reply to this

Re: Cake Talk

Is anything known about flows on manifolds which are locally modelled on reproducing kernel hilbert spaces?

I am not even sure I know what this question means! In which way is your flow supposed to be modeled on a Hilbert space?

Posted by: urs on February 22, 2007 5:05 PM | Permalink | Reply to this

Re: Cake Talk

I just meant flows on infinite-dimensional Riemannian manifolds, but where the charts are maps to subspaces of a RKHS. The kind of model used in nonparametric statistics has this form.

I know next to nothing about infinite-dimensional geometry. I see there is such a thing as a Hilbert manifold, and that a Hilbert space can be given the structure of a Riemannian manifold.

Posted by: David Corfield on February 22, 2007 5:22 PM | Permalink | Reply to this

single-celled inference via Bayes; Re: Cake Talk

Bacteria use Bayes’ Law in their evolved “inference” behavior.

How cells deal with uncertainty

Researchers at McGill University have found that cells respond to their ever-changing environment in a way that mimics the optimal mathematical approach to doing so, also known as Bayes’ rule; an application of probability theory. Their findings are published in the April 17 issue of PNAS, the Proceedings of the National Academy of Sciences.

Biology is seeing a re-birth, said Dr. Peter Swain, an assistant professor in the Department of Physiology and a Canada Research Chair in Systems Biology, as more researchers are thinking about the cell using schemes that we know work from engineering and computer science.

The study was carried out at McGill Centre for Nonlinear Dynamics in Physiology and Medicine (CND). Eric Libby, PhD candidate at the CND and lead author on the paper, Dr. Ted Perkins, assistant professor in the School of Computer Science, and Dr. Swain simulated data on a biochemical response mechanism in a strain of E. coli bacteria.

The ideal mathematical model and the simulation meshed perfectly with Bayes rule,” remarked Swain. The bacterias collection of genes and proteins that responded to changing environmental conditions acted as a successful Bayesian inference module, which takes noisy, uncertain information and interprets what it means for the cell.

There are many known schemes for inference that exist in mathematics. This study suggests that cells may have evolved to incorporate the most efficient decision-making abilities into their biochemical pathways.

Quick, accurate cell responses to signals are necessary for survival. When we sense danger, our bodies can tell if the signal is real and trigger the production of adrenaline immediately. However, modeling the effects of a signal on one part of a cell, even in isolation from body tissues and organs, is complicated. With many drugs, we dont know how they work or exactly what they are targeting in a cell noted Swain. He explained that further study of inference modules could allow us to model more sophisticated cellular behavior, which could one day lead to computerized drug experiments and trials.

Source: McGill University

Posted by: Jonathan Vos Post on April 21, 2007 8:16 PM | Permalink | Reply to this

Post a New Comment