Untangling Lariats: Subgradient Following of Variationally Penalized Objectives
This work addresses smoothing and sparsity challenges in temporal and multivariate data analysis, presenting incremental improvements by extending existing variational penalty frameworks.
The paper tackles the problem of optimizing convex objectives with variational penalties for smoothing sequences, deriving known algorithms like fused lasso and isotonic regression as special cases, and introduces new penalties and efficient solvers for high-order filtering and multivariate problems.
We describe an apparatus for subgradient-following of the optimum of convex problems with variational penalties. In this setting, we receive a sequence $y_i,\ldots,y_n$ and seek a smooth sequence $x_1,\ldots,x_n$. The smooth sequence needs to attain the minimum Bregman divergence to an input sequence with additive variational penalties in the general form of $\sum_i{}g_i(x_{i+1}-x_i)$. We derive known algorithms such as the fused lasso and isotonic regression as special cases of our approach. Our approach also facilitates new variational penalties such as non-smooth barrier functions. We then derive a novel lattice-based procedure for subgradient following of variational penalties characterized through the output of arbitrary convolutional filters. This paradigm yields efficient solvers for high-order filtering problems of temporal sequences in which sparse discrete derivatives such as acceleration and jerk are desirable. We also introduce and analyze new multivariate problems in which $\mathbf{x}_i,\mathbf{y}_i\in\mathbb{R}^d$ with variational penalties that depend on $\|\mathbf{x}_{i+1}-\mathbf{x}_i\|$. The norms we consider are $\ell_2$ and $\ell_\infty$ which promote group sparsity.