Dataset Information

Deconvolution When Classifying Noisy Data Involving Transformations.

ABSTRACT: In the present study, we consider the problem of classifying spatial data distorted by a linear transformation or convolution and contaminated by additive random noise. In this setting, we show that classifier performance can be improved if we carefully invert the data before the classifier is applied. However, the inverse transformation is not constructed so as to recover the original signal, and in fact, we show that taking the latter approach is generally inadvisable. We introduce a fully data-driven procedure based on cross-validation, and use several classifiers to illustrate numerical properties of our approach. Theoretical arguments are given in support of our claims. Our procedure is applied to data generated by light detection and ranging (Lidar) technology, where we improve on earlier approaches to classifying aerosols. This article has supplementary materials online.

SUBMITTER: Carroll R

PROVIDER: S-EPMC3630802 | biostudies-literature | 2012 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deconvolution When Classifying Noisy Data Involving Transformations.

Carroll Raymond R Delaigle Aurore A Hall Peter P

Journal of the American Statistical Association 20120901 499

In the present study, we consider the problem of classifying spatial data distorted by a linear transformation or convolution and contaminated by additive random noise. In this setting, we show that classifier performance can be improved if we carefully invert the data before the classifier is applied. However, the inverse transformation is not constructed so as to recover the original signal, and in fact, we show that taking the latter approach is generally inadvisable. We introduce a fully dat ...[more]

PMID: 23606778

Similar Datasets

Project description:BackgroundA suitable disease classification is essential for individualized therapy in patients with Crohn's disease (CD). Although a potential mechanistic classification of colon-involving and non-colon-involving disease was suggested by recent genetic and microbiota studies, the clinical implication has seldom been investigated. We aimed to explore the association of this colonic-based classification with clinical outcomes in patients with CD compared with the Montreal classification.MethodsThis was a retrospective study of CD patients from a tertiary referral center. Patients were categorized into colon-involving and non-colon-involving disease, and according to the Montreal classification. Clinico-demographic data, medications, and surgeries were compared between the two classifications. The primary outcome was the need for major abdominal surgery.ResultsOf 934 patients, those with colonic involvement had an earlier median (interquartile range) age of onset [23.0 (17.0-30.0) versus 26.0 (19.0-35.0) years, p = 0.001], higher frequency of perianal lesions (31.2% versus 14.5%, p < 0.001) and extraintestinal manifestations (21.8% versus 14.5%, p = 0.010), but lower frequency of stricture (B2) (16.3% versus 24.0%, p = 0.005), than those with non-colon-involving disease. Colon-involving disease was a protective factor against major abdominal surgery [hazard ratio, 0.689; 95% confidence interval (CI), 0.481-0.985; p = 0.041]. However, patients with colon-involving CD were more prone to steroids [odds ratio (OR), 1.793; 95% CI, 1.206-2.666; p = 0.004] and azathioprine/6-mercaptopurine (AZA/6-MP) treatment (OR, 1.732; 95% CI, 1.103-2.719; p = 0.017) than were patients with non-colon-involving disease. The Montreal classification was not predictive of surgery or steroids and AZA/6-MP treatment.ConclusionThis study supports the rationale for disease classification based on the involvement of colon. This new classification of CD is a better predictor of clinical outcomes than the Montreal classification.

Project description:Most connectivity metrics in neuroimaging research reduce multivariate activity patterns in regions-of-interests (ROIs) to one dimension, which leads to a loss of information. Importantly, it prevents us from investigating the transformations between patterns in different ROIs. Here, we applied linear estimation theory in order to robustly estimate the linear transformations between multivariate fMRI patterns with a cross-validated ridge regression approach. We used three functional connectivity metrics that describe different features of these voxel-by-voxel mappings: goodness-of-fit, sparsity and pattern deformation. The goodness-of-fit describes the degree to which the patterns in an input region can be described as a linear transformation of patterns in an output region. The sparsity metric, which relies on a Monte Carlo procedure, was introduced in order to test whether the transformation mostly consists of one-to-one mappings between voxels in different regions. Furthermore, we defined a metric for pattern deformation, i.e. the degree to which the transformation rotates or rescales the input patterns. As a proof of concept, we applied these metrics to an event-related fMRI data set consisting of four subjects that has been used in previous studies. We focused on the transformations from early visual cortex (EVC) to inferior temporal cortex (ITC), fusiform face area (FFA) and parahippocampal place area (PPA). Our results suggest that the estimated linear mappings explain a significant amount of response variance in the three output ROIs. The transformation from EVC to ITC shows the highest goodness-of-fit, and those from EVC to FFA and PPA show the expected preference for faces and places as well as animate and inanimate objects, respectively. The pattern transformations are sparse, but sparsity is lower than would have been expected for one-to-one mappings, thus suggesting the presence of one-to-few voxel mappings. The mappings are also characterised by different levels of pattern deformations, thus indicating that the transformations differentially amplify or dampen certain dimensions of the input patterns. While our results are only based on a small number of subjects, they show that our pattern transformation metrics can describe novel aspects of multivariate functional connectivity in neuroimaging data.

Dataset Information

Deconvolution When Classifying Noisy Data Involving Transformations.

Publications

Deconvolution When Classifying Noisy Data Involving Transformations.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets