Learning computational methods for partial differential equations from the web

Abstract

A course has been developed to learn computational methods from the web1 and has been tested with postgraduate students from re- mote universities. Short video conferences or video recordings provide an overview and introduce more detailed studies with numerical experiments in Java-powered course notes. This enables every participant to work at his own pace and to develop his intuition for finite differences, finite elements, Fourier, Monte-Carlo and Lagrangian methods. Assignments are carried out in a regular web browser and are compiled into web pages where the students explain with their own words, equations and programs how to derive, implement and run computational schemes. Our experience shows that the technology is rapidly acquired from templates, using practical examples for the advection, diffusion, Black-Scholes, Burger, Korteweg-DeVries and Schrödinger equations.