The fourth conundrum is here! Let’s dive back into the world of modeling with this weeks data surrounding properties.
Attached is information regarding just over twenty thousand properties in Victoria, Canada. The data* includes various metrics such as location, elevation, the height of the property, extent of the property below ground and the use of the property.
Can you use all this information to build a model that predicts the use of the property?** Be careful: with over 40% of the properties having the same use it would be easy to bias your model towards over-predicting that particular use type.
*For the avoidance of doubt: 'ActualUse' is the column with the target variable.
Hi - replying here as well as in PM for others visibility, this is a know bug that we are working on and will get fixed ASAP!