Data Science Balancing Act
Data science is a continuous balancing act. A data scientist needs to be able to examine and validate detailed data processing steps and model parameters, but also remember the big picture to make sure that the model they are building is a viable solution to business problems. They need to be able to understand complex mathematics, statistics, and computational topics but explain them in a way that someone who doesn’t can understand.
A data scientist needs to iteratively test new modeling methods and fine tune parameters but also know how to stop when the results are good enough, and needs to identify appropriate trade-offs between processing time and model complexity, precision and recall, and predictive and explanatory power. They should be self-motivated and independently curious, but must be able to collaborate and ask for help, because data science is too broad for anyone to know everything. Most importantly, a data scientist needs to continually reflect and seek feedback on whether they are striking the right balance, and be willing to shift accordingly.
By: Jen Heider