Varying Coefficient Linear Discriminant Analysis for Dynamic Data
This provides an incremental improvement for statisticians and machine learning practitioners working with heterogeneous dynamic classification data.
The paper tackles classification of dynamic data by developing a varying coefficient linear discriminant analysis model where the discriminant direction varies with an exposure variable, proposing a computationally efficient B-spline-based estimation method that achieves optimal or near-optimal estimation error rates in different dimensional regimes.
Linear discriminant analysis (LDA) is an important classification tool in statistics and machine learning. This paper investigates the varying coefficient LDA model for dynamic data, with Bayes' discriminant direction being a function of some exposure variable to address the heterogeneity. We propose a new least-square estimation method based on the B-spline approximation. The data-driven discriminant procedure is more computationally efficient than the dynamic linear programming rule \citep{jiang2020dynamic}. We also establish the convergence rates for the corresponding estimation error bound and the excess misclassification risk. The estimation error in $L_2$ distance is optimal for the low-dimensional regime and is near optimal for the high-dimensional regime. Numerical experiments on synthetic data and real data both corroborate the superiority of our proposed classification method.