The details of the imputation method for each feature with missing values can be summarized in this table:
Feature | Imputation method | Details |
|---|---|---|
| product_group | most frequent | applied for all families |
| end_use | most frequent | applied for all families |
| gbu_customer_seg | most frequent | applied for all families |
| manual_region | most frequent | applied for all families |
| lip2 | most frequent | applied only for Alkoxylates |
| historical_unit_price_coalesce_ratio_on_12 | constant | value equal to 1 for all families |
| historical_sales_coalesce_ratio_on_12 | constant | value equal to 1 for all families |
| historical_unit_price_ratio_3_on_12_month | constant | value equal to 1 for all families |
| n_competitors | constant | applied for Sulfosuccinate_Sulfosuccinamate & Sulfosuccinates_Healthcare value equal to 1 |
| COMPONENT_ratio | mean | applied for all families |
| IMPURITY_ratio | mean | applied for all families |
| SOLVENT_ratio | mean | applied for all families |
| n_components | mean | applied for all families |