... It's not just a Kaggle competition!
To evaluate the technical skills of our Data Science candidates, we kindly ask you to solve the challenge in the link below. This is a demand forecasting problem where you must build a model able to forecast weekly sales for different departments, stores, and retailers.
- Your resolution must be developed and delivered on a Kaggle notebook;
- Please, build your solution in an organized and structured way, because we want to understand what steps you took to develop your model;
- We also expect to read a text containing your solution justifying the decisions made and highlighting the main conclusions (example: why you decided to plot a graph, explain what conclusions you can extract from it; or why you are choosing a particular model, make it clear which was the criterion for choice). It's super important that our team can follow your way of thinking without following the lines of code!
- Code quality is also very important, so keep in mind that well-written code warms our hearts!
- Super tip: a clear and well-structured rationale is more important than a model with very high performance! Also, we often need to explain our reasoning to a non-technical audience, so it's worth exercising your ability to explain complex concepts simply!
Once you finish the challenge, please follow the instructions on this form.
Important!
- Remember to make your Kaggle notebook public so we can access it;
- Be aware to present your solution if called for a technical interview!
Good luck and have fun,
Ze´s Data Science Team