Data Science tasks are graded manually, so the hiring manager or recruiter will need to assign a score after a candidate submits their test. Here are some things to look for when grading a data science task submission:
- Are they able to familiarise themselves with a new dataset through meaningful exploration and visualizations of the data
- Are they able to wrangle large amounts of data and restructure it to fit their modeling purposes. In this case, several datasets must be joined and transformed in order to get the variables on net bike stocks in stations
- Choose a type of model from a large variety of available modeling techniques. A clear justification of the choice of a certain type of model over another shows a good understanding of the theoretical and practical implications
- Show that their results are generalizable and thus useful for the business. In order to do this some form of test set must be set up (see below for more on this point) and the results should be clearly reported/visualized
After looking over the candidate's solution, the reviewer can assign a score by clicking on the task score's edit button. If you have any questions, email us at firstname.lastname@example.org!