📅 DISC 212 – Class Scheduler Optimization Using ILP

📌 Project Overview

An academic project for DISC 212 focused on optimizing university class schedules using Integer Linear Programming (ILP). Built with Python and Pulp, the system considers constraints like class timings, venue capacity, and space availability to generate the most efficient schedule possible.

I developed Model-2, which integrates a weighted constraint approach to achieve higher venue usage and schedule balance. Our team compared this model against the university's real schedule and an alternative model (Model-1) designed by another member. Model-2 consistently delivered superior results.

🎯 Objectives

• Maximize space utilization across available time slots and venues
• Account for constraints: seating capacity, venue availability, class strength, and scheduling patterns
• Generate venue-specific insights including average usage and class distributions
• Compare scheduling outcomes across models and real-world data

🔍 Key Features

• Optimized scheduling using pulp_cbc_cmd solver
• Comparative visualizations (bar charts, heatmaps) for MW/TR course spread and venue performance
• Venue-level utilization stats showing mean classes and average capacity usage
• Evaluation of scheduling quality across three models

🛠️ Technologies Used

• Python
• Pulp (ILP solver)
• Pandas & NumPy
• Matplotlib
• Jupyter Notebook

📊 Future Enhancements

• Include instructor preferences and teaching patterns
• Add event types (lectures, labs, night classes, etc.)
• Integrate support for venue pre-bookings and non-teaching usage

📂 Repository Structure

├── RO                                 # University's schedule stats
├── model_1                            # Alternative model by another team member
├── model_2                            # Optimized model using ILP (Usama's contribution) with MW/TR schedule and stats
├── evaluation.ipynb                   # Visual comparisons of all models and venue/course stats
├── dataset (courses.csv, venues.csv)  # Input dataset (venue, courses, schedule info) extracted from university's website
├── README.md                          # Documentation and project overview