What are the common pitfalls in Data Science projects?

Common pitfalls in Data Science projects include overfitting, where a model acts particularly well on the training data and poorly on new data, and leakage of data, where information outside the training dataset clumsily leaks into the model. Low data quality and lack of objectives contribute to project failure. The models approval by a data scientist is required rigorously to make sure that the understanding of a business problem is existent.