Details

Autor(en) / Beteiligte

• Orujlu, Vusal

Titel

Implementation of ANN in Software Effort Estimation: Boundary Value Effort Forecast :A novel Artificial Neural Networks Model to Improve the Accuracy of Effort Estimation in Software Development Projects

Ort / Verlag

ProQuest Dissertations & Theses

Erscheinungsjahr

2021

Quelle

ProQuest Dissertations & Theses A&I

Beschreibungen/Notizen

• Software Development consistently accommodates a variety of unstable scenarios. Good planning always stands behind well-defined requirements. Hence, the consistency of the effort estimation plays a special role in the traditional Business-Consumer relationship.While the proposed models may provide high accuracy in predicting specific data sets, it’s still difficult for IT specialists/organizations to find the best method for evaluating certain functionalities. The challenge of the project; initiated programming language, project infrastructure, and/or staff experimentation are just a few of the reasons that lead to inequality in these terms.Conceptually, the planned work going to explicate the main correlations. It will contain historical background - as to how was the industrial lifecycle before pre-processing progress/what was the necessity for them to exist, as well as modern usage area of BPM and Project Management – like how managers and owners’ moves are intending to keep the consumer’s satisfaction in higher level while increasing the revenue.Taking the most failure causes of projects into consideration, the research will capture some components of Software Project Management to clarify developed approaches and their advantages and/or disadvantages. The study may also lead somehow to the Business Process Management to see the alignments of required tasks in a rigorous way.The research is generally intending to define the key features of the Project Effort Estimation as usage of the datasets, evaluating the architectures, etc. The investigation also aims to find effective causes of poor effort estimation and analyze how those improvable points may be developed to ensure a highly accurate Artificial Neural Networks model.