Revolutionizing Inefficient Scheduling Systems by Adapting Offline Programming Techniques to Function in an Online Algorithm within a Python Application

Abstract

Modern scheduling mechanisms fail to accommodate real-world time management constraints, often wasting time and preventing optimal usage. To solve this problem, two approaches may be used: Offline and Online Algorithms. Offline algorithms receive full input before creating output, while online algorithms receive input incrementally and make decisions without full data. While many studies cover offline algorithms, there is a lack of knowledge about online ones. We designed a program that analyzes key factors, including time slot popularity and reservation history, to make optimal choices. Users enter their desired time range for an activity, and the algorithm determines the best room. The efficiency of this room selection is measured by the greatest potential to meet the needs of future users. Preliminary data suggests that in comparison to both a random room selection as well as the default offline algorithm using linear programming (considered to be an optimal output as it is given the full data), the online algorithm developed performs around 250% better than a random one but only 75% as well as the offline one (based on number of users scheduled in total). With larger amounts of historical data, it is highly likely that this algorithm would be able to outperform current results, but in its current form, it still serves as a valuable scheduling tool and a foundation for future research.