Nomograms for Linear Models
Требуемые условия завершения
Set up the following workflow:
- Load the data on Human development index.
- In Select Columns, set the Continent as a target variable.
- Construct Naive Bayesian model and view it in a histogram.
Now, explore the following questions.
- Which are the most important features for deciding whether some country is in Europe or not?
- Which variables can potentially offer the strongest evidence against being in Europe?
- Connect Table to Select Columns, and the connect Table's output to the Nomogram (without removing or changing any existing connections). In Table, choose Russian Federation. This will set the Nomogram so that it will show data for Russia. In Nomogram, set "Rank by" to Absolute importance. Now you can see which are the most European and the most un-European features of Russia.
Последнее изменение: пятница, 11 марта 2022, 17:39