2019/04/08 by Alexander Greaves-Tunnell, Zaid Harchaoui, Zaïd Harchaoui +2
Computer Science · Engineering · Mathematics · #Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Music and Audio Processing #Neural Networks and Applications #Sound (cs.SD) #Time Series Analysis and Forecasting #cs.LG #cs.SD #eess.AS #electronic engineering #information engineering #stat.ML
paper · pdf · doi:10.48550/arxiv.1904.03834
29 pages; expanded supplement, added details in background and methods per reviewer feedback, included additional references
openalex publication_date 2019/04/08 · arxiv created 2019/06/07 · arxiv updated 2019/06/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Representation and learning of long-range dependencies is a central challenge confronted in modern applications of machine learning to sequence data. Yet despite the prominence of this issue, the basic problem of measuring long-range dependence, either in a given data source or as represented in a trained deep model, remains largely limited to heuristic tools. We contribute a statistical framework for investigating long-range dependence in current applications of deep sequence modeling, drawing on the well-developed theory of long memory stochastic processes. This framework yields testable implications concerning the relationship between long memory in real-world data and its learned representation in a deep learning architecture, which are explored through a semiparametric framework adapted to the high-dimensional setting.