The Data Science / Data Analytics / CRISP-DM Cycle
I have data lying around. How do I start - and continue - if I want to build a predictive model based on these data? Well, it is not really like following a cooking recipes with precise steps. It is more like adjusting the steps here and there, going back and starting again with different parameters or maybe even more drastically anew with different algorithms.
This video explains this general iterative process.