Abstract: Data poisoning is an attack on machine learning models that involves inserting a sample into the training data sets that does not influence the model during training but does affect it ...