Multivariate calibration on heterogeneous samples

Data heterogeneity has become a challenging problem in modern data analysis. Classic statistical modeling methods, which assume the data are independent and identically distributed, often show unsatisfactory perfor-mance on heterogeneous data. This work is motivated by a multivariate calibration problem from a soil char-acterization study, where the samples were collected from five different locations. Newly proposed and existing signal regression models are applied to the multivariate calibration problem, where the models are adapted to handle such spatially clustered structure. When compared to a variety of other methods, e.g. kernel ridge regression, random forests, and partial least squares, we find that our newly proposed varying-coefficient signal regression model is highly competitive, often out-performing the other methods, in terms of external prediction error.

Automated summary

Data heterogeneity poses a challenge in modern data analysis, with traditional statistical modeling methods struggling to perform well on such data. This study addresses a multivariate calibration problem in soil characterization, proposing a varying-coefficient signal regression model that outperforms other methods in external prediction error.