Скачать или смотреть #20: Scikit-learn 17: Preprocessing 17: Univariate feature imputation: SimpleImputer

The video discusses the intuition for missing values in a dataset. Next, it discusses the code for univariate feature imputation in Scikit-learn in Python.

Timeline
(Python 3.8)

00:00 - Outline of video
00:49 - What are missing values? Numeric
01:40 - What is imputation?
03:55 - What are missing values? Categorical
05:07 - What is Univariate imputation?
06:07 - What is CSR and CSC matrix?
08:16 - Code snippet (--- CORRECTION ---: the line "strategy =..." needs to be commented)
10:13 - Open Jupyter notebook
11:10 - Create an array with missing values using np.nan
11:47 - SimpleImputer: missing_values=np.nan, strategy='mean'
14:20 - Convert an array to a CSC matrix
16:24 - SimpleImputer: with CSC matrix: missing_values=-1, strategy='mean'
19:15 - Create a DataFrame with missing values
20:42 - SimpleImputer: with Dataframe: missing_values=np.nan, strategy='most_frequent'
21:35 - Ending notes