Improving the performance of effort estimation in terms of function point analysis by balancing datasets
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F70883521%3A28140%2F23%3A63570239" target="_blank" >RIV/70883521:28140/23:63570239 - isvavai.cz</a>
Result on the web
<a href="https://link.springer.com/chapter/10.1007/978-3-031-21435-6_60" target="_blank" >https://link.springer.com/chapter/10.1007/978-3-031-21435-6_60</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1007/978-3-031-21435-6_60" target="_blank" >10.1007/978-3-031-21435-6_60</a>
Alternative languages
Result language
angličtina
Original language name
Improving the performance of effort estimation in terms of function point analysis by balancing datasets
Original language description
This research proposes an approach to improve the performance of effort estimation based on the balancing of each group for categorical variables. The proposed model is based on function point analysis, Industry Sector, and deep learning. The Pytorch library is used to build the deep learning model with the dataset ISBSG (release 2020). The accuracy of our model is compared with that of the Adj-Effort approach. We adopt the prediction level at 0.3, Mean Absolute Error, Mean Balanced Relative Error, Mean Inverted Balanced Relative Error, and Standardised Accuracy as criteria for validation. The findings demonstrate that our proposed model outweighs the unbalanced and Adj-Effort approaches. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
—
Continuities
S - Specificky vyzkum na vysokych skolach
Others
Publication year
2023
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Lecture Notes in Networks and Systems
ISBN
978-3-031-21434-9
ISSN
23673370
e-ISSN
2367-3389
Number of pages
10
Pages from-to
705-714
Publisher name
Springer Science and Business Media Deutschland GmbH
Place of publication
Berlín
Event location
on-line
Event date
Oct 10, 2022
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—