Leverage Fix and Optimize method to learn decision variables to solve Flexible stochastic resource constrained project scheduling problem
Budget: €250 – €750 EUR
Python+Machine Learning+ Project Scheduling + Mathematical Knowledge
The Resource-Constrained Project Scheduling Problem (RCPSP) is a challenging combinatorial optimization problem with numerous practical applications in project management, manufacturing, and other domains. The objective is to determine an optimal schedule for a set of activities, taking into account various constraints such as activity precedence relationships, resource availability, and project completion time.
Efficiently solving the RCPSP has significant implications for real-world scenarios. By finding optimal or near-optimal schedules, project managers can effectively allocate resources, minimize project duration, and enhance project performance. However, due to the problem's complexity and the presence of multiple constraints, finding optimal solutions for large-scale instances remains a computationally demanding task.
The traditional approach to solving the RCPSP involves the application of mathematical programming techniques and heuristic algorithms. While these methods have made significant progress, there is still a need for novel approaches that can improve solution quality and computational efficiency.
This research proposes the integration of Fix and Optimize methods with RCPSP to tackle the existing challenges and further advance the state of the art in solving this problem. The Fix and Optimize paradigm combines mathematical programming models with machine learning techniques to learn decision variable patterns and iteratively refine solutions, resulting in improved solution quality and reduced computational effort.
The primary motivation for this research is twofold. Firstly, by leveraging Fix and Optimize methods, we aim to enhance the efficiency and effectiveness of solving the RCPSP, enabling practitioners to handle larger problem instances and achieve more optimal schedules. Secondly, the integration of machine learning techniques with optimization models opens up opportunities for learning decision-making patterns and generating valuable insights for project scheduling.
To support this research, a comprehensive dataset of 25 problem instances for 10 real activities and 25 problem instances for 30 real activities has been collected. This dataset provides a foundation for initial research and enables empirical evaluations of the proposed methods.
By addressing the existing limitations in RCPSP solution approaches and exploring the potential of Fix and Optimize methods, this research aims to contribute to both academia and industry. The findings of this study can help project managers make more informed decisions, optimize resource allocation, and improve project outcomes.
Overall, this research aims to provide valuable insights into the application of Fix and Optimize methods for solving the RCPSP, advancing the understanding of optimization techniques in project scheduling and contributing to the broader field of combinatorial optimization.
The Resource-Constrained Project Scheduling Problem (RCPSP) is a challenging combinatorial optimization problem with numerous practical applications in project management, manufacturing, and other domains. The objective is to determine an optimal schedule for a set of activities, taking into account various constraints such as activity precedence relationships, resource availability, and project completion time.
Efficiently solving the RCPSP has significant implications for real-world scenarios. By finding optimal or near-optimal schedules, project managers can effectively allocate resources, minimize project duration, and enhance project performance. However, due to the problem's complexity and the presence of multiple constraints, finding optimal solutions for large-scale instances remains a computationally demanding task.
The traditional approach to solving the RCPSP involves the application of mathematical programming techniques and heuristic algorithms. While these methods have made significant progress, there is still a need for novel approaches that can improve solution quality and computational efficiency.
This research proposes the integration of Fix and Optimize methods with RCPSP to tackle the existing challenges and further advance the state of the art in solving this problem. The Fix and Optimize paradigm combines mathematical programming models with machine learning techniques to learn decision variable patterns and iteratively refine solutions, resulting in improved solution quality and reduced computational effort.
The primary motivation for this research is twofold. Firstly, by leveraging Fix and Optimize methods, we aim to enhance the efficiency and effectiveness of solving the RCPSP, enabling practitioners to handle larger problem instances and achieve more optimal schedules. Secondly, the integration of machine learning techniques with optimization models opens up opportunities for learning decision-making patterns and generating valuable insights for project scheduling.
To support this research, a comprehensive dataset of 25 problem instances for 10 real activities and 25 problem instances for 30 real activities has been collected. This dataset provides a foundation for initial research and enables empirical evaluations of the proposed methods.
By addressing the existing limitations in RCPSP solution approaches and exploring the potential of Fix and Optimize methods, this research aims to contribute to both academia and industry. The findings of this study can help project managers make more informed decisions, optimize resource allocation, and improve project outcomes.
Overall, this research aims to provide valuable insights into the application of Fix and Optimize methods for solving the RCPSP, advancing the understanding of optimization techniques in project scheduling and contributing to the broader field of combinatorial optimization.