Simple predictions for supermarket visitors (10 hrs to do)
The dataset contains daily visitor data from a shopping attraction. For some features, no description is available, so they may be dummy features (also called 1-hot encoded variables, binary indicators,...), numerical features, etc., but not categorical ones. For the following features we provide additional information:
school holiday: 0 = no school holiday 1 = school holiday
bank holiday: 0 = no bank holiday 1 = bank holiday
Additionally, daily weather data for the location of the shopping attraction is provided.
The task is to predict the column called 'label' for the test set. We will measure the prediction error using the (root-)mean-squared error ((R)MSE) metric. You may use any programming language and freely available libraries to solve this task.
Hint: An appropriate baseline will give an RMSE of approx. 500.