Math 658-600 Final Exam Review
The final exam will be given on December 11, 2012, from 8-10 am in our usual classroom. We will skip the material on the Radon transform. The test will cover sections 9.1-9.4 in the text. In addition, it will cover the Notes on Scattered-Data Radial Function Interpolation, the
Notes on Spherical Harmonics, and the material on rotations that we discussed in class. The test will have 5 to 7 questions, some with
multiple parts; questions will involve problems or propositions
similar to homework problems (assignments 6 to 9) or those done in class. You will have a choice of which problems to work. Half of the test will be over distributions and tempered distributions.
Distributions and tempered distributions
- Spaces of test functions
- The space $\mathscr D(\mathbb R^n)$
- Definition, multi-index notation, examples (§9.1)
- Convergence in $\mathscr D$. (Notes for 10/24/12 and part (ii) of the definition on p. 306 in the text).
- Operations mapping $\mathscr D$ continuously to itself differentiation, multiplication by a $C^\infty$ function, convolutions, and so on. (§9.2)
- Schwartz space $\mathscr S(\mathbb R^n)$
- Definition and notation (§9.4)
- Semi-norm and (equivalent) metric space topologies (See HW 6 and notes, 10/26/12)
- Operations continuously mapping $\mathscr S$ to itself differentiation, multiplication by polynomials, Fourier transform, and so on.
- Be able to sketch a proof of this theorem: The Fourier transform on $\mathscr S$ is a bijection continuous in the topology of $\mathscr S$. (See text and notes, 10/29/12.)
- The spaces $\mathscr D'$ and $\mathscr S'$
- Distributions, $\mathscr D'$
- Definition and notation, examples, $\delta$ function, and convergence of a sequence of distributions
- Operations on $\mathscr D'$ differentiation
multiplication by a $C^\infty$ function, and so on. (§9.1 - §9.3)
- Tempered distributions, $\mathscr S'$
- Know the definition as a continuous linear functional on $\mathscr S$ (see notes, 10/31/12 and 11/2/12), and also know notation examples, and convergence of sequences of tempered distributions (§9.3 and HW 7., 8.)
- Know the definition of a finite order tempered distribution and the know and be able to use continuity test: A linear functional defined on $\mathscr S$ is in $F\in \mathscr S'$ if and only if $F$ is of finite order. (Notes, 11/2/12.)
- Operations on $\mathscr S'$ differentiation,
multiplication by polynomials, taking Fourier transforms, etc.
- Be able to find Fourier transforms of tempered distributions. (See HW 6 and HW 7.)
Reproducing kernel Hilbert spaces (RKHS)
- Definition, simple properties, and examples (notes, 11/7/12 and 11/9/12)
- Be able to the existence of the associated reproducing kernel
- Interpolation and connection with least-squares minimization problems (notes, 11/7/12 and 11/9/12)
- Positive definite and conditionally positive definite kernels
- Associated RKHS and inner product (notes, 11/12/12)
Scattered-data interpolation via RBFs
- Scattered data interpolation problems. (See the Notes on
Scattered-Data Radial Function Interpolation. Theorems, propositions, etc. all refer to these notes.))
- Interpolation without polynomial reproduction
- Interpolation with polynomial reproduction
- Polynomials and unisolvent sets
- Strictly positive definite and conditionally positive definite functions (see Definition 2.1)
- Be able to prove Proposition 2.2.
- Completely monotonic functions (Widder's theorem)
- Radial functions that are positive definite for all $\mathbb R^n$'s. Schoenberg's theorem be able use it to give examples
- Order $m$ RBFs Theorem 2.5. See examples in §III.
- Native spaces and the power function (notes, 11/19/12 and HW 8)
Spherical harmonics
- Laplace-Beltrami for $\mathbb S^2$. (See the Notes on Spherical Harmonics.)
- Spherical coordinates and how to handle singularities at the poles.
- The eigenvalue problem
- Be able to get the bound in Theorem 2.1
- Be able to sketch a proof of Proposition 2.2
- The spherical harmonics. (You do not need to memorize the formulas in in sections 2 and 3.)
- Know how to use raising and lowering operators to go from one spherical harmonic to another
- Know that when homogeneous harmonic polynomials on $\mathbb R^3$ are restricted to the sphere $\mathbb S^2$ they are spherical harmonics.
Rotations in $\mathbb R^3$
- Definition and properties of 3×3 rotation matrices
- Be able to show that the eigenvalues of a rotation matrix $R$ satisfy $|\lambda|=1$ and that $\lambda =1 $ is always an eigenvalue
- SKIP the Euler angles
- Be able to give a geometric derivation of the formula $R\mathbf x = \mathbf x + \sin(\theta)\mathbf \omega \times \mathbf x+ (1-\cos(\theta))\mathbf \omega\times (\mathbf \omega \times \mathbf x)$. (Notes, 12/3/12.)
- SU(2) and SO(3)
- Be able to express $U\in SU(2)$ in terms of quaternions and to use it to obtain the connection between $SU(2)$ and the sphere $\mathbb S^3$.
- Be able to do a problem similar to problem 7 in HW 9. (You won't be asked to do any long calculations.)
- Know the connection between $SU(2)$ and $SO(3)$ and the one between $SO(3)$ and the projective space $\mathbb RP^3$.
Updated 12/7/2012 (fjn).