Search for: All records

Subdata selection from big data is an active area of research that facilitates inferences based on big data with limited computational expense. For linear regression models, the optimal design-inspired Information-Based Optimal Subdata Selection (IBOSS) method is a computationally efficient method for selecting subdata that has excellent statistical properties. But the method can only be used if the subdata size, k, is at last twice the number of regression variables, p. In addition, even when $$k\ge 2p$$, under the assumption of effect sparsity, one can expect to obtain subdata with better statistical properties by trying to focus on active variables. Inspired by recent efforts to extend the IBOSS method to situations with a large number of variables p, we introduce a method called Combining Lasso And Subdata Selection (CLASS) that, as shown, improves on other proposed methods in terms of variable selection and building a predictive model based on subdata when the full data size n is very large and the number of variables p is large. In terms of computational expense, CLASS is more expensive than recent competitors for moderately large values of n, but the roles reverse under effect sparsity for extremely large values of n. 

Aspergillus flavus is an agriculturally important fungus that causes ear rot of maize and produces aflatoxins, of which B 1 is the most carcinogenic naturally-produced compound. In the US, the management of aflatoxins includes the deployment of biological control agents that comprise two nonaflatoxigenic A . flavus strains, either Afla-Guard (member of lineage IB) or AF36 (lineage IC). We used genotyping-by-sequencing to examine the influence of both biocontrol agents on native populations of A . flavus in cornfields in Texas, North Carolina, Arkansas, and Indiana. This study examined up to 27,529 single-nucleotide polymorphisms (SNPs) in a total of 815 A . flavus isolates, and 353 genome-wide haplotypes sampled before biocontrol application, three months after biocontrol application, and up to three years after initial application. Here, we report that the two distinct A . flavus evolutionary lineages IB and IC differ significantly in their frequency distributions across states. We provide evidence of increased unidirectional gene flow from lineage IB into IC, inferred to be due to the applied Afla-Guard biocontrol strain. Genetic exchange and recombination of biocontrol strains with native strains was detected in as little as three months after biocontrol application and up to one and three years later. There was limited inter-lineage migration in the untreated fields. These findings suggest that biocontrol products that include strains from lineage IB offer the greatest potential for sustained reductions in aflatoxin levels over several years. This knowledge has important implications for developing new biocontrol strategies. 

To study brain activity, by measuring changes associated with the blood flow in the brain, functional magnetic resonance imaging techniques are employed. The design problem in event-related functional magnetic resonance imaging studies is to find the best sequence of stimuli to be shown to subjects for precise estimation of the brain activity. Previous analytical studies concerning optimal functional magnetic resonance imaging designs often assume a simplified model with independent errors over time. Optimal designs under this model are called g-lag orthogonal designs. Recently, it has been observed that g-lag orthogonal designs also perform well under simplified models with auto-regressive error structures. However, these models do not include drift. We investigate the performance of g-lag orthogonal designs for models that incorporate drift parameters. Identifying g-lag orthogonal designs that perform best in the presence of a drift is important because a drift is typically assumed for the analysis of event-related functional magnetic resonance imaging data. 

Abstract Magic squares have been extremely useful and popular in combinatorics and statistics. One generalization of magic squares ismagic rectangleswhich are useful for designing experiments in statistics. A necessary and sufficient condition for the existence of magic rectangles restricts the number of rows and columns to be either both odd or both even. In this paper, we generalize magic rectangles to even by oddnearly magic rectangles. We also prove necessary and sufficient conditions for the existence of a nearly magic rectangle, and construct one for each parameter set for which they exist.