Distributional Representation of Cyclic Alternating Patterns for A-Phase Classification in Sleep EEG

This article describes a detailed methodology for the A-phase classification of cyclic alternating patterns (CAPs) present in sleep electroencephalography (EEG). CAPs are valuable EEG marker instability and represent an important pattern with which to analyze additional characteristics processes, manifestations have been linked some specific conditions. CAP phase detection not commonly carried out routinely due time attention this problem requires (and if present, labels user-dependent, visually evaluated, hand-made); thus, automatic tool solve is presented. The experiments were using distributional representation data obtained from Sleep Database. For purpose, symbolization was performed one-dimensional symbolic aggregate approximation (1d-SAX), followed by vectorization trained Doc2Vec model final ten classic machine learning models two separate validation strategies. best results support vector classifier radial basis kernel. hold-out validation, F1 Score 0.7651; stratified 10-fold cross-validation, 0.7611 ± 0.0133. illustrates that proposed suitable classification.