The article discusses the utilization of Sequential Neural Networks (SNNs) in Functional Data Analysis (FDA). SNNs are neural networks designed to process sequence data, a crucial aspect of functional data. The authors suggest that using SNNs in FDA applications can potentially overcome the limitations of traditional FDA methods, allowing for greater scalability, flexibility, and improved analytical performance. The study includes a comparative analysis of SNNs against popular FDA regression models, demonstrating their effectiveness in both numerical experiments and real-world data analysis.
Publication date: 6 Nov 2023
Project Page: arXiv:2311.01875v1 [cs.LG] 3 Nov 2023
Paper: https://arxiv.org/pdf/2311.01875
