Measure Selection for Functional Linear Model
By: Su I Iao, Hans-Georg Müller
Potential Business Impact:
Helps analyze health data better with new math.
Advancements in modern science have led to an increased prevalence of functional data, which are usually viewed as elements of the space of square-integrable functions $L^2$. Core methods in functional data analysis, such as functional principal component analysis, are typically grounded in the Hilbert structure of $L^2$ and rely on inner products based on integrals with respect to the Lebesgue measure over a fixed domain. A more flexible framework is proposed, where the measure can be arbitrary, allowing natural extensions to unbounded domains and prompting the question of optimal measure choice. Specifically, a novel functional linear model is introduced that incorporates a data-adaptive choice of the measure that defines the space, alongside an enhanced function principal component analysis. Selecting a good measure can improve the model's predictive performance, especially when the underlying processes are not well-represented when adopting the default Lebesgue measure. Simulations, as well as applications to COVID-19 data and the National Health and Nutrition Examination Survey data, show that the proposed approach consistently outperforms the conventional functional linear model.
Similar Papers
Clustering of functional data prone to complex heteroscedastic measurement error
Methodology
Finds hidden groups in messy data.
Robust Functional Logistic Regression
Methodology
Makes computer predictions more accurate, even with bad data.
Functional principal component analysis for functional data with detection limits
Methodology
Fixes data errors when numbers are too small or big.