Deterministic Decoupling of Global Features and its Application to Data Analysis
This addresses the problem of feature interdependence for researchers and practitioners in data analysis, offering an incremental improvement over classical descriptors.
The paper tackles the problem of decoupling global features in data analysis by introducing a deterministic method based on transformations along feature gradients, which yields a quasi-analytic solution for orthokurtosis. The result is a consistent and significant performance improvement in regression and classification tasks compared to using non-decoupled descriptors.
We introduce a method for deterministic decoupling of global features and show its applicability to improve data analysis performance, as well as to open new venues for feature transfer. We propose a new formalism that is based on defining transformations on submanifolds, by following trajectories along the features gradients. Through these transformations we define a normalization that, we demonstrate, allows for decoupling differentiable features. By applying this to sampling moments, we obtain a quasi-analytic solution for the orthokurtosis, a normalized version of the kurtosis that is not just decoupled from mean and variance, but also from skewness. We apply this method in the original data domain and at the output of a filter bank to regression and classification problems based on global descriptors, obtaining a consistent and significant improvement in performance as compared to using classical (non-decoupled) descriptors.