Use the available climate variables, measured from different weather stations, to predict the Instant Air Temperature at a given time.

The data was taken hourly from different weather stations in South America. Therefore information regarding the time, date and location of the station is also available.

The data is incomplete, some of it was omitted or not recorded, meaning teams will have make use of data-cleaning methods and procedures available through use of the R programming language.

You can submit up to 2 predictions a day. The public score is the RMSE of 30% of the test data. An example for a submission is presented in the data section.

As you progress through the course, you will learn more methods for data cleaning, exploratory analysis and model building - try coming back to the data from week to week and use what you have learned! Your team could win first place!