If you have participated in the Software Entwicklungs Praktikum (SEP), you are likely familiar with the student assignment system currently in use. The process begins with supervisors specifying projects and their capacity limits. Students submit their preferences for projects and potential team partners, along with detailed skill profiles. Supervisors then evaluate these entries, endorsing or vetoing students based on their suitability. They can also issue explicit rejections for students with whom there have been past issues. However, the decision-making process heavily favors student preferences, often overlooking other important factors.
Your task is to design a more sophisticated algorithm that refines this process, incorporating a fully functional system demonstrable to the SEP organizer. This system should improve on multiple fronts:
Refine the data collection mechanism to capture comprehensive, algorithm-specific information from students and supervisors. Consider including:
- Detailed preferences for projects and team partners.
- Comprehensive skill sets and proficiencies.
- Supervisor requirements and preferences concerning student capabilities and project needs.
- Specifications on minimum and maximum student counts per project.
- Thresholds for potential project cancellation.
Evaluate the essential data points to ensure robust assignment outcomes without imposing undue data entry burdens. There are various frameworks such as Streamlit that allow you to quickly build interactive web applications for data collection and visualization. To make things easy, you could also start by collecting the data in simple CSV files and only develop the web application later, once you know what data you need.
Construct a mathematical model to precisely outline your optimization objectives and constraints:
- Objectives:
- Maximize the student placement rate within projects.
- Prioritize assignments based on nuanced student preferences, possibly by allowing students to score rather than rank projects.
- Facilitate team formations based on preferred partnerships.
- Integrate supervisor feedback to align student assignments with project skill demands.
- Target ideal project team sizes to prevent overfilling or underfilling.
- Constraints:
- Enforce specific student capacity ranges for each project, with provisions for cancelling projects that do not meet minimum interest levels.
- Implement skill-based placements, such as ensuring a balance of front-end and back-end developers or meeting minimum skill prerequisites for projects.
Note that you do not need to implement all of these objectives and constraints. Just think about which ones are most important and feasible to implement. A too complex model might be to unpredictable regarding what it will consider good solution, and it also might be too slow for optimization. If you can come up with a model that is both simple and effective, that would be great.
Develop and fine-tune an algorithm that efficiently solves this enhanced model, leveraging the latest in optimization technologies and libraries.
- Produce outputs in a machine-readable format that can be easily processed by the SEP organizer.
- Establish benchmarks to gauge the effectiveness of your algorithm in creating optimal group formations based on the identified objectives and constraints. You could for example design 20 typical student profiles and just duplicate them by some estimated numbers to get a larger dataset. You can do the same for the projects.
- Ensure the algorithm's scalability to accommodate scenarios involving up to 1000 students and 100 projects.
Initiate the project by developing a foundational model that addresses the core objectives and constraints - primarily, maximizing the assignment of students to their preferred projects while adhering to the group size limitations. Incrementally incorporate more sophisticated objectives and constraints, such as skill matching and supervisor preferences. This phased approach will help manage the complexity of the system, allowing for iterative enhancements and adjustments based on testing and feedback.
This project offers a significant opportunity to apply your optimization skills to a tangible challenge, with the potential to substantially improve the SEP assignment process. We wish you the best in your endeavors to innovate and refine this system.
Effectively communicate the value of your solution in the final presentation. Aim to demonstrate convincingly to the organizer that your algorithm delivers superior assignments compared to the existing system. Ideally, include a live demonstration of your system to showcase its functionality and user interface in real time. If the optimization process for larger datasets requires extended time, prepare a video demonstration that highlights key features and successful outputs. This approach will help illustrate the practical benefits and efficiency of your solution, making a compelling case for its adoption.
- May 7th: Project Kickoff
- Bi-Weekly: Check-ins with us. You can decide the weeks, but we would like to see some progress every two weeks.
- July 8-12th: Final Presentation