I am currently working on a supervised learning to predict project success from project features. Currently, the data shows 99% of the projects were successfully completed which results in a huge class imbalance. Does anyone know if there is a way of identifying projects that received re-allocated funds. e.g. is there a specific donor_id that I can look for.
Potentially, I would like to exclude such projects from my experiments for reasons I cannot get into at the moment. Any help would be greatly appreciated.
Thank you for helping tackle this question!