Data science, being a relatively new field, has no go-to version control standard. In this post I discuss data science cycles, its unique versioning strategy, and some possible solutions. Data Science Cycles Data science is different from traditional software development, especially so for stages such as exploratory data analysis, feature engineering, machine learning model training & validation.