Hitting-time problems with moving boundaries
We are happy to present Martin Nilsson, a Mathematical physicist. Associate Professor (Sw. “Docent”) at RISE. His research focuses on machine learning and AI, but he has worked multidisciplinary in, e.g., signal processing, mechatronics, robotics, computer science, and biophysics.
Date and time: 2 February 2021, 3 pm – 4 pm
Speaker: Martin Nilsson
Title: Hitting-time problems with moving boundaries
Zoom: https://kth-se.zoom.us/j/67432682790?pwd=dVgzbjRSbUVFT2FOYTByYlZrTU9BUT09
Meeting ID: 674 3268 2790
Password: DF2020
Watch the recorded presentation:
Abstract: Suppose that you borrow money to buy a house. The bank requires the house as a security for the loan. What is the probability that you will lose the house, i.e., that the market value of the house falls below your debt before you have paid off the loan? This is an example of a hitting-time problem, common in natural sciences and finance. It is often considered difficult due to the complication that the debt changes over time as you pay off the loan.
In the first part of the talk, I will introduce the field by presenting a few different examples of hitting-time problems. In the second part, I will briefly sketch an approach that can be used to solve a wide class of hitting-time problems with moving boundaries very efficiently. Whereas the second part is more technical than the first, some elementary previous encounter with the heat equation should be perfectly sufficient for understanding the details.
Bio: Martin Nilsson, is a Mathematical physicist. Associate Professor (Sw. “Docent”) at RISE. His research focuses on machine learning and AI, but he has worked multidisciplinary in, e.g., signal processing, mechatronics, robotics, computer science, and biophysics. He is especially interested in neurobiophysics and the brain’s tipping us off how knowledge can be efficiently represented and processed in computers.