February 26, 2007

Alan Hammond (NYU):   The uncertainty principle  

Abstract:

The uncertainty principle states that if a function is supported mainly in a box of sidelength $\delta$, and its Fourier transform mainly in a box of sidelength $\rho$, then $\delta \rho \geq 1$. The uncertainty principle suggests a technique for bounding the number of eigenvalues of an elliptic operator: instead of volume counting, which can lead to wild over-estimates, we count the number of unit boxes that can be symplectically embedded in phase space, thereby approximately diagonalizing the operator. The embeddings must also satisfy certain smoothness requirements. The talk is a presentation of a part of "the uncertainty principle", C.Fefferman,  Bull. Amer. Math. Soc. (1983) no 2, 129-206.


back to the seminar homepage