% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@INPROCEEDINGS{Miliou:1025669,
author = {Yu, Jinyang and Hamdan, Sami and Sasse, Leonard and
Morrison, Abigail and Patil, Kaustubh R.},
editor = {Miliou, Ioanna and Piatkowski, Nico and Papapetrou,
Panagiotis},
title = {{E}mpirical {C}omparison {B}etween {C}ross-{V}alidation
and {M}utation-{V}alidation in {M}odel {S}election},
volume = {14642},
address = {Cham},
publisher = {Springer Nature Switzerland},
reportid = {FZJ-2024-03058},
isbn = {978-3-031-58555-5 (print)},
series = {Lecture Notes in Computer Science},
pages = {56 - 67},
year = {2024},
comment = {Advances in Intelligent Data Analysis XXII / Miliou, Ioanna
(Editor) [https://orcid.org/0000-0002-1357-1967] ; Cham :
Springer Nature Switzerland, 2024, Chapter 5 ; ISSN:
0302-9743=1611-3349 ; ISBN:
978-3-031-58555-5=978-3-031-58553-1 ;
doi:10.1007/978-3-031-58553-1},
booktitle = {Advances in Intelligent Data Analysis
XXII / Miliou, Ioanna (Editor)
[https://orcid.org/0000-0002-1357-1967]
; Cham : Springer Nature Switzerland,
2024, Chapter 5 ; ISSN:
0302-9743=1611-3349 ; ISBN:
978-3-031-58555-5=978-3-031-58553-1 ;
doi:10.1007/978-3-031-58553-1},
abstract = {Mutation validation (MV) is a recently proposed approach
for model selection, garnering significant interest due to
its unique characteristics and potential benefits compared
to the widely used cross-validation (CV) method. In this
study, we empirically compared MV and k-fold CV using
benchmark and real-world datasets. By employing Bayesian
tests, we compared generalization estimates yielding three
posterior probabilities: practical equivalence, CV
superiority, and MV superiority. We also evaluated the
differences in the capacity of the selected models and
computational efficiency. We found that both MV and CV
select models with practically equivalent generalization
performance across various machine learning algorithms and
the majority of benchmark datasets. MV exhibited advantages
in terms of selecting simpler models and lower computational
costs. However, in some cases MV selected overly simplistic
models leading to underfitting and showed instability in
hyperparameter selection. These limitations of MV became
more evident in the evaluation of a real-world
neuroscientific task of predicting sex at birth using brain
functional connectivity.},
month = {Apr},
date = {2024-04-24},
organization = {22nd International Symposium on
Intelligent Data Analysis, IDA 2024,
Proceedings, Part II, Stockholm
(Sweden), 24 Apr 2024 - 26 Apr 2024},
cin = {INM-7 / IAS-6},
cid = {I:(DE-Juel1)INM-7-20090406 / I:(DE-Juel1)IAS-6-20130828},
pnm = {5253 - Neuroimaging (POF4-525) / 5254 - Neuroscientific
Data Analytics and AI (POF4-525)},
pid = {G:(DE-HGF)POF4-5253 / G:(DE-HGF)POF4-5254},
typ = {PUB:(DE-HGF)8 / PUB:(DE-HGF)7},
UT = {WOS:001295920900005},
doi = {10.1007/978-3-031-58553-1_5},
url = {https://juser.fz-juelich.de/record/1025669},
}