KNN imputation
Datasets always have features which are correlated. Hence, it becomes important to consider them as a factor for imputing missing values. Machine learning models use features in the DataFrame to find correlations and patterns and predict a selected feature.
One of the simplest and most efficient models is the K Nearest Neighbors. It finds 'K' points most similar to the existing data points to impute missing values.
In this exercise, the diabetes
DataFrame has already been loaded for you. Use the fancyimpute
package to impute the missing values in the diabetes
DataFrame.
This exercise is part of the course
Dealing with Missing Data in Python
Exercise instructions
- Import
KNN
fromfancyimpute
. - Copy
diabetes
todiabetes_knn_imputed
. - Create a
KNN()
object and assign it toknn_imputer
. - Impute the
diabetes_knn_imputed
DataFrame.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Import KNN from fancyimpute
___
# Copy diabetes to diabetes_knn_imputed
diabetes_knn_imputed = ___
# Initialize KNN
knn_imputer = ___
# Impute using fit_tranform on diabetes_knn_imputed
diabetes_knn_imputed.iloc[:, :] = ___