Multiscale Analysis for Intensity and Density Estimation Rebecca Willett’s MS Defense Thanks to Rob Nowak, Mike Orchard, Don Johnson, and Rich Baraniuk Eric Kolaczyk and Tycho Hoogland Poisson and Multinomial Processes Why study Poisson Processes? Astrophysics Network analysis Medical Imaging Multiresolution Analysis Examining data at different resolutions (e.g., seeing the forest, the trees, the leaves, or the dew) yields different information about the structure of the data. Multiresolution analysis is effective because it sees the forest (the overall structure of the data) without losing sight of the trees (data singularities) Beyond Wavelets Multiresolution analysis is a powerful tool, but what about… Non-Gaussian problems? Edges? Image Edges? Nongaussian noise? Inverse problems? Piecewise polynomial- and plateletbased methods address these issues. Computational Harmonic Analysis I. Define Class of Functions to Model Signal A. Piecewise Polynomials B. Platelets II. Derive basis or other representation III. Threshold or prune small coefficients IV. Demonstrate near-optimality Approximating Besov Functions with Piecewise Polynomials Error Decay Rate : Od 2r d-r d r Discrete Error Decay : O N N N 3 2 Approximation with Platelets Consider approximating this image: 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 50 100 150 200 250 E.g. Haar analysis 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 50 100 150 Terms = 2068, Params = 2068 200 250 Wedgelets Original Image Haar Wavelet Partition Wedgelet Partition E.g. Haar analysis with wedgelets 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 50 100 150 Terms = 1164, Params = 1164 200 250 E.g. Platelets 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 50 100 Terms = 510, Params = 774 150 200 250 Error Decay Platelet Approximation Theory b,2 Holder (Cb ) a,1 Holder (Ca ) Error decay rates: • Fourier: O(m-1/2) • Wavelets: O(m-1) • Wedgelets: O(m-1) • Platelets: O(mmin(a,b)) 0,0 0,0 0,0 1,0 1,1 1,0 1,0 2,0 3,0 4,0 2,1 3,1 4,1 1,1 4,2 3,2 4,3 4,4 2,2 3,3 4,5 1,1 4,6 3,4 4,7 4,8 4,9 2,3 3,5 4,10 4,11 3,6 4,12 4,13 3,7 4,14 4,15 A Piecewise Constant Tree A Piecewise Linear Tree Maximum Penalized Likelihood Estimation Goal: Maximize the penalized likelihood L (μ) log p( x | μ) {# θ} where p( x | μ) is the likelihood is a smoothing parameter {# θ} is the number of parameters in θ So the MPLE is ˆ arg max L (μ(θ )) θ θ ˆ) ˆ μ(θ μ The Algorithm Data Const Estimate Wedge Estimate Platelet Estimate Wedged Platelet Estimate Inherit from finer scale Algorithm in Action Penalty Parameter Penalty parameter balances between fidelity to the data (likelihood) and complexity (penalty). 1 0.8 Bias Choose c log N 2 0.6 Different values 0.4 0.2 0 0 2 4 6 Variance = 0 Estimate is MLE: μˆ x Estimate is a constant: μˆ x 8 10 Penalization Bowl 0 10 Ave. Counts/Pixel = 1 Ave. Counts/Pixel = 10 Ave. Counts/Pixel = 100 -1 Mean MSE 10 -2 10 -3 10 -4 10 -2 10 -1 0 10 10 b / log(# of photon counts) 1 10 Density Estimation - Blocks Density Estimation - Heavisine Density Estimation - Bumps Density Estimation Simulation Medical Imaging Results Inverse Problems Goal: estimate m from observations x ~ Poisson(Pm) EM algorithm (Nowak and Kolaczyk, ’00): E Step : Q(m , m ) Emi [L (m ) | x] i M - Step : max Q(m , m ) i m Confocal Microscopy: An Inverse Problem Platelet Performance 2.6 MLE (min error level shown in dots) Piecewise Constant Platelet 2.4 2.2 Error 2 1.8 1.6 1.4 1.2 2 4 6 8 Iteration 10 12 14 Confocal Microscopy: Real Data Hellinger Loss H pm , pm ' 2 pm ( x ) pm ' ( x ) • Upper bound for affinity A(p, q) p( x )q( x ) 1/ 2 (like squared error) • Relates expected error to Lp approximation bounds 2 Bound on Hellinger Risk (follows from Li & Barron ’99) e EH p ,p pen( m ' ) If m ' 2 m mˆ 1 , then risk of mˆ satisfies min KL pm ,pm' pen(m' ) m ' KL distance Approximation error Estimation error Bounding the KL • We can show: 1 n 2N 2 KL pμ , pμ' μ μ' 2 N N c • Recall approximation result: d-r d r O N N N 3 2 μ - μ' 2 • Choose optimal d 2 Near-optimal Risk 1 O N 2r 2r 1 2r 2 2 r 1 1 H2 (p , p ) O log (N) μˆ μ N N • Maximum risk within logarithmic factor of minimum risk • Penalty structure effective: c log N Conclusions CHA with Piecewise Polynomials or Platelets • • • • Effectively describe Poisson or multinomial data Strong approximation capabilites Fast MPLE algorithms for estimation and reconstruction Near-optimal characteristics Major Contributions • Risk analysis for piecewise polynomials • Platelet representations and approximation theory Future Work • Shift-invariant methods • Fast algorithms for wedgelets and platelets • Risk Analysis for platelets Approximation Theory Results Consider the class of functions f ( x, y ) f1 ( x, y ) I{ y H ( x )} f 2 ( x, y ) (1 I{ y H ( x )} ) b ,2 a ,1 where f i Holder (C b ), and H Holder (Ca ). If fˆ is the m - term, J - scale, resolution platelet approximat ion, so 2 m 2 , then J f fˆ L2 Ka , b m min( a , b ) . Why don’t we just find the MLE? - λi e λ p(x | λ ) x i! i xi i xi log( p(x | λ )) 1 0 λ i i λˆ x MPLE Algorithm (1D) Multiscale Likelihood Factorization Probabilistic analogue to orthonormal wavelet decomposition Parameters wavelet coefficients Allow MPLE framework, where penalization based on complexity of underlying partition Poisson Processes • Goal: Estimate spatially varying function, (i,j), from observations of Poisson random variables x(i,j) with intensities (i,j) • MLE of would simply equal x. We will use complexity regularization to yield smoother estimate. Accurate Model Complexity Regularization Parsimonious Model Penalty for each constant region results in fewer splits Bigger penalty for each polynomial or platelet region more degrees of freedom, so more efficient to store constant if likely Astronomical Imaging