Development of a Machine Learning Based Predictive Model for Event Management

Event management is crucial for the success of any event, as a well-managed event leads to high attendee and client satisfaction. Traditional event management systems emphasize efficient tracking of user data and electronic booking but often lack the ability to provide clients with pre-visualization experience of their event venue based on their specifications. This study addresses this gap by proposing an event management system capable of generating wedding event design styles based on user inputs using a deep learning model. The system leverages a dataset of diverse wedding decoration images, undergoing preprocessing, feature extraction, and image labeling to facilitate the deep learning model, Stable Diffusion, in understanding and predicting new decoration styles from user inputs. Stable Diffusion, an image generative deep learning model, learns from past datasets to create images from text prompts. The system features a web application with a Graphical User Interface (GUI) for user input, generating images based on these inputs. After 500 epochs, the system achieved an accuracy of 0.9765 using precision metrics. The system offers a predictive modeling approach to provide clients with pre-visualization experience, creating decoration images tailored to their preferences. Future work includes enhancing the model's performance by training with larger datasets and expanding the application to other event types such as birthdays and funerals.