Graph-Based Regularization for Regression Problems with Alignment and Highly Correlated Designs
نویسندگان
چکیده
منابع مشابه
Graph-based regularization for regression problems with highly-correlated designs
Abstract: Sparse models for high-dimensional linear regression and machine learning have received substantial attention over the past two decades. Model selection, or determining which features or covariates are the best explanatory variables, is critical to the interpretability of a learned model. Much of the current literature assumes that covariates are only mildly correlated. However, in mo...
متن کاملOptimal designs in regression with correlated errors.
This paper discusses the problem of determining optimal designs for regression models, when the observations are dependent and taken on an interval. A complete solution of this challenging optimal design problem is given for a broad class of regression models and covariance kernels. We propose a class of estimators which are only slightly more complicated than the ordinary least-squares estimat...
متن کاملMetamorphosis problems for graph designs
A λ-fold G-design of order n is a pair (X, B), where B is a collection of edge disjoint copies of the simple graph G which partitions the edge set of λK n (the undirected complete graph with n vertices) with vertex set X. Let (X, B) be a G-design and H be a subgraph of G. For each block b ∈ B, partition b into copies of H and G \ H and place the copy of H in B(H) and the edges belonging to the ...
متن کاملRobust Designs for Approximate Regression Models with Correlated Errors
We summarize recent ...ndings on the construction of designs, to be used in experiments for which regression is the anticipated method of analysis. The designs are to be robust against departures from the assumed linear response and/or departures from the assumption of uncorrelated errors. Included is a discussion of an “in...nitesimal” approach to design. In this approach one aims to minimize ...
متن کاملOptimal designs for comparing regression models with correlated observations
We consider the problem of efficient statistical inference for comparing two regression curves estimated from two samples of dependent measurements. Based on a representation of the best pair of linear unbiased estimators in continuous time models as a stochastic integral, an efficient pair of linear unbiased estimators with corresponding optimal designs for finite sample size is constructed. T...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: SIAM Journal on Mathematics of Data Science
سال: 2020
ISSN: 2577-0187
DOI: 10.1137/19m1287365