Computer Science > Computational Engineering, Finance, and Science
[Submitted on 15 Mar 2021 (v1), revised 20 Nov 2021 (this version, v5), latest version 27 Oct 2022 (v8)]
Title:Online Learning with Radial Basis Function Networks
View PDFAbstract:We investigate the benefits of feature selection, nonlinear modelling and online learning when forecasting in financial time series. We consider the sequential and continual learning sub-genres of online learning. The experiments we conduct show that there is a benefit to online transfer learning, beyond the sequential updating of recursive least-squares models. We show that feature representation transfer via radial basis function networks, which make use of clustering algorithms to construct a kernel Gram matrix, are more beneficial than treating each training vector as separate basis functions, as occurs with kernel Ridge regression. We also demonstrate quantitative procedures to determine the very structure of the networks. Finally, we conduct experiments on the log returns of financial time series and show that these online transfer learning models are able to outperform a random walk baseline, whereas the offline learning models struggle to do so.
Submission history
From: Gabriel Borrageiro Mr [view email][v1] Mon, 15 Mar 2021 14:39:40 UTC (194 KB)
[v2] Wed, 23 Jun 2021 08:26:21 UTC (1,096 KB)
[v3] Mon, 8 Nov 2021 07:14:08 UTC (1,095 KB)
[v4] Mon, 15 Nov 2021 09:57:33 UTC (2,081 KB)
[v5] Sat, 20 Nov 2021 10:37:35 UTC (1,513 KB)
[v6] Fri, 24 Dec 2021 11:41:31 UTC (1,511 KB)
[v7] Sat, 21 May 2022 12:29:19 UTC (1,544 KB)
[v8] Thu, 27 Oct 2022 08:31:30 UTC (190 KB)
Current browse context:
cs.CE
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.