Authors : Rakibul Islam, Md Anwar Hosen, Md Borhan Uddin, Saima Tasnim, Sabrina Shamim Moushi, Suny Md Ashraf Khan, Khandakar Rabbi Ahmed
Publication date : 2025/11/20
Conference : 2025 9th International Conference on Computational System and Information Technology for Sustainable Solutions (CSITSS)
Pages : 1-6
Publisher : IEEE
Description : Innovation in project management requires advanced analysis techniques that can model multiple unrelated project attributes and their interdependencies. Traditional project management techniques cannot represent the nonlinear interactions among attributes and relational project attributes within portfolios. In this work, we discuss HPGNet, a novel hybrid artificial intelligence system comprising Multilayer Perceptron MLP and Graph Convolutional Network GCN for learning high-dimensional tabular project information and graph-structured relational information in parallel. Using an openly available dataset of IT and construction projects, HPGNet constructs similarity graphs to capture relationships between projects and integrates them with dense feature embeddings to accurately predict multi-class risks and optimally allocate resources. Our approach outperforms single-modality baselines by a landslide …
Total citations : Cited by 14
Scholar articles :
R Islam, MA Hosen, MB Uddin, S Tasnim, SS Moushi… – 2025 9th International Conference on Computational …, 2025