Kontaktujte nás | Jazyk: čeština English
Název: | Software cost estimation using neural networks |
Autor: | Ramaekers, Robin; Šilhavý, Radek; Šilhavý, Petr |
Typ dokumentu: | Článek ve sborníku (English) |
Zdrojový dok.: | Lecture Notes in Networks and Systems. 2023, vol. 722 LNNS, p. 831-847 |
ISSN: | 2367-3370 (Sherpa/RoMEO, JCR) |
ISBN: | 978-303135310-9 |
DOI: | https://doi.org/10.1007/978-3-031-35311-6_77 |
Abstrakt: | Software Cost Estimation (SCE) is one of the most vital parts when starting a new software engineering project; it helps with allocating resources, managing risks, making informed decisions, and stands in correlation with the success or the failure of a project. Because Software Cost Estimation (SCE) is prone to human bias, solutions started being researched with the aid of Artificial Intelligence (AI) and Machine Learning (ML). This paper will investigate the importance of Software Cost Estimation (SCE). Further, the existing taxonomies and methodologies regarding using neural networks with Software Cost estimation will be compared (COCOMO, GEHO-ANN, OLCE, and -ANN-NEAT). This will be done using evaluation metrics such as RMSE, MMRE, PRED, MAE, etc. After, further research is proposed on why using Deep Reinforcement Learning (DRL) could be very beneficial for developing Software Cost Prediction Models. This technique combines Deep Learning (DL) and Machine Learning (ML) and can solve complex tasks with many variables and a rapidly developing environment. |
Plný text: | https://link.springer.com/chapter/10.1007/978-3-031-35311-6_77 |
Zobrazit celý záznam |
Soubory | Velikost | Formát | Zobrazit |
---|---|---|---|
K tomuto záznamu nejsou připojeny žádné soubory. |