Feature Importance#

Description:

Machine Learning Time Series model to predict future values. This can be based on daily, monthly, or hourly values.

Function:

Feature_Importance(table = string, value_col = string, measure_cols = list,
n_estimators = integer, random_state = integer, new_table_name = string)

Parameters:

  • Table: Table name on which to perform the function

  • Value Column: Column to measure the feature importance on

  • Measure Columns: Columns test feature importance

  • Number of Estimators: Number of trees: more trees can be more accurate but take longer (pick a number between 10 & 500)

  • Random State: Random seed: this will allow you to replicate the model (pick a number between 1 & 100)

  • New Table Name: Name for the new table

Example:

Feature_Importance(table = Budget, value_col = "Revenue", measure_cols = ["Business Unit", "Employee Size", "Geo"],
n_estimators = 42, random_state = 99, new_table_name = "Feature Imp." )