Grey Market Prediction

This thesis aims to use machine learning, specifically classification algorithms, to predict which of the company’s products could end up on a grey market. The goal was to develop a tool to enhance the probability of finding such sales and potentially allowing them to act before this violation occurred. The results of this model will used by the specific company for which it was developed but could be also beneficial as a guide to other companies facing the same problem with their products ending on a grey market. This thesis describes all stages of the project from data extraction, data cleaning and manipulation, feature engineering, model training, prediction, and creation of the visual from the results. In the end, the use of the current outputs is explained together with how the use of the model could be extended in the future.