---
_id: '1203'
abstract:
- lang: eng
text: " “Principal Component Analysis” (PCA) is an established
linear technique for dimensionality reduction. It performs an orthonormal transformation
to replace possibly correlated variables with a smaller set of linearly independent
variables, the so-called principal components, which capture a large portion of
the data variance. The problem of finding the optimal number of principal components
has been widely studied for offline PCA. However, when working with streaming
data, the optimal number changes continuously. This requires to update both the
principal components and the dimensionality in every timestep. While the continuous
update of the principal components is widely studied, the available algorithms
for dimensionality adjustment are limited to an increment of one in neural network-based
and incremental PCA. Therefore, existing approaches cannot account for abrupt
changes in the presented data. The contribution of this work is to enable in neural
network-based PCA the continuous dimensionality adjustment by an arbitrary number
without the necessity to learn all principal components. A novel algorithm is
presented that utilizes several PCA characteristics to adaptivly update the optimal
number of principal components for neural network-based PCA. A precise estimation
of the required dimensionality reduces the computational effort while ensuring
that the desired amount of variance is kept. The computational complexity of the
proposed algorithm is investigated and it is benchmarked in an experimental study
against other neural network-based and incremental PCA approaches where it produces
highly competitive results.\r\n "
article_number: e0248896
author:
- first_name: Nico
full_name: Migenda, Nico
id: '218473'
last_name: Migenda
orcid: 0000-0002-7223-1735
- first_name: Ralf
full_name: Möller, Ralf
last_name: Möller
- first_name: Wolfram
full_name: Schenck, Wolfram
id: '224375'
last_name: Schenck
orcid: 0000-0003-3300-2048
orcid_put_code_url: https://api.orcid.org/v2.0/0000-0003-3300-2048/work/94914468
citation:
alphadin: 'Migenda, Nico ; Möller,
Ralf ; Schenck, Wolfram:
Adaptive dimensionality reduction for neural network-based online principal component
analysis. In: PLOS ONE Bd. 16, Public Library of Science (PLoS) (2021),
Nr. 3'
ama: Migenda N, Möller R, Schenck W. Adaptive dimensionality reduction for neural
network-based online principal component analysis. PLOS ONE. 2021;16(3).
doi:10.1371/journal.pone.0248896
apa: Migenda, N., Möller, R., & Schenck, W. (2021). Adaptive dimensionality
reduction for neural network-based online principal component analysis. PLOS
ONE, 16(3). https://doi.org/10.1371/journal.pone.0248896
bibtex: '@article{Migenda_Möller_Schenck_2021, title={Adaptive dimensionality reduction
for neural network-based online principal component analysis}, volume={16}, DOI={10.1371/journal.pone.0248896},
number={3e0248896}, journal={PLOS ONE}, publisher={Public Library of Science (PLoS)},
author={Migenda, Nico and Möller, Ralf and Schenck, Wolfram}, year={2021} }'
chicago: Migenda, Nico, Ralf Möller, and Wolfram Schenck. “Adaptive Dimensionality
Reduction for Neural Network-Based Online Principal Component Analysis.” PLOS
ONE 16, no. 3 (2021). https://doi.org/10.1371/journal.pone.0248896.
ieee: N. Migenda, R. Möller, and W. Schenck, “Adaptive dimensionality reduction
for neural network-based online principal component analysis,” PLOS ONE,
vol. 16, no. 3, 2021.
mla: Migenda, Nico, et al. “Adaptive Dimensionality Reduction for Neural Network-Based
Online Principal Component Analysis.” PLOS ONE, vol. 16, no. 3, e0248896,
Public Library of Science (PLoS), 2021, doi:10.1371/journal.pone.0248896.
short: N. Migenda, R. Möller, W. Schenck, PLOS ONE 16 (2021).
date_created: 2021-06-03T19:35:49Z
date_updated: 2023-05-10T13:40:20Z
doi: 10.1371/journal.pone.0248896
intvolume: ' 16'
issue: '3'
language:
- iso: eng
publication: PLOS ONE
publication_identifier:
eissn:
- 1932-6203
publication_status: published
publisher: Public Library of Science (PLoS)
status: public
title: Adaptive dimensionality reduction for neural network-based online principal
component analysis
type: journal_article
user_id: '249199'
volume: 16
year: '2021'
...