Enhancing Success Prediction In Crowdfunding Platform Using LIME-Explainable AI Method

This study focuses on utilizing various machine learning techniques, ranging from supervised models to ensemble boosting methods, to predict the success of crowdfunding campaigns. The project aims to make these models more transparent and interpretable during decision making by using explainable AI (XAI) methods such as LIME. The dataset used for training these models is obtained from the well-known crowdfunding platform, Kickstarter. By leveraging XAI techniques, the project gains insights into the model's functionality, identifying important features that impact predictions positively or negatively. This understanding enables the refinement of machine learning models, leading to more precise predictions. In summary, the project comprises an effective combination of machine learning and explainable AI, demonstrating how these technologies can work together to produce highly precise, accurate, and interpretable success prediction models.