Fourier-based parametrization of convolutional neural networks for robust time series forecasting


Krstanovic, Sascha ; Paulheim, Heiko


DOI: https://doi.org/10.1007/978-3-030-33778-0_39
URL: https://link.springer.com/chapter/10.1007/978-3-03...
Additional URL: http://www.heikopaulheim.com/docs/ds2019.pdf
Document Type: Conference or workshop publication
Year of publication: 2019
Book title: Discovery Science 22nd International Conference, DS 2019, Split, Croatia, October 28–30, 2019, Proceedings
The title of a journal, publication series: Lecture Notes in Computer Science
Volume: 11828
Page range: 522-532
Conference title: DS 2019
Location of the conference venue: Split, Croatia
Date of the conference: 28.-30.10.2019
Author/Publisher of the book
(only the first ones mentioned)
:
Kralj Novak, Petra
Place of publication: Cham
Publishing house: Springer Nature
ISBN: 978-3-030-33777-3 , 978-3-030-33778-0
Publication language: English
Institution: School of Business Informatics and Mathematics > Web Data Mining (Paulheim 2018-)
Subject: 004 Computer science, internet
Abstract: Classical statistical models for time series forecasting most often make a number of assumptions about the data at hand, there-with, requiring intensive manual preprocessing steps prior to modeling. As a consequence, it is very challenging to come up with a more generic forecasting framework. Extensive hyperparameter optimization and ensemble architectures are common strategies to tackle this problem, however, this comes at the cost of high computational complexity. Instead of optimizing hyperparameters by training multiple models, we propose a method to estimate optimal hyperparameters directly from the characteristics of the time series at hand. To that end, we use Convolutional Neural Networks (CNNs) for time series forecasting and determine a part of the network layout based on the time series’ Fourier coefficients. Our approach significantly reduces the amount of required model configuration time and shows competitive performance on time series data across various domains. A comparison to popular, state of the art forecasting algorithms reveals further improvements in runtime and practicability.

Dieser Eintrag ist Teil der Universitätsbibliographie.




+ Citation Example and Export

Krstanovic, Sascha ; Paulheim, Heiko ORCID: 0000-0003-4386-8195 Fourier-based parametrization of convolutional neural networks for robust time series forecasting. Kralj Novak, Petra Lecture Notes in Computer Science 11828 522-532 In: Discovery Science 22nd International Conference, DS 2019, Split, Croatia, October 28–30, 2019, Proceedings (2019) Cham DS 2019 (Split, Croatia) [Conference or workshop publication]


+ Search Authors in

BASE: Krstanovic, Sascha ; Paulheim, Heiko

Google Scholar: Krstanovic, Sascha ; Paulheim, Heiko

ORCID: Krstanovic, Sascha ; Paulheim, Heiko ORCID: 0000-0003-4386-8195

+ Page Views

Hits per month over past year

Detailed information



You have found an error? Please let us know about your desired correction here: E-Mail


Actions (login required)

Show item Show item