填補連續型遺漏值

Feature Engineering for Machine Learning in Python

Robert O'Callaghan

Director of Data Science, Ordergroove

刪除遺漏值

  • 測試集含遺漏值時，不能直接刪除整列
Feature Engineering for Machine Learning in Python

還能怎麼做？

  • 類別欄：以最常見值取代，或用標示遺漏的字串（如「None」）
  • 數值欄：以合適的數值取代
Feature Engineering for Machine Learning in Python

集中趨勢量數

  • 平均數（mean）
  • 中位數（median）
Feature Engineering for Machine Learning in Python

計算集中趨勢量數

print(df['ConvertedSalary'].mean())
print(df['ConvertedSalary'].median())

92565.16992481203
55562.0
Feature Engineering for Machine Learning in Python

填補遺漏值

df['ConvertedSalary'] = df['ConvertedSalary'].fillna(
    df['ConvertedSalary'].mean()
)

df['ConvertedSalary'] = df['ConvertedSalary']\
                         .astype('int64')
Feature Engineering for Machine Learning in Python

四捨五入

df['ConvertedSalary'] = df['ConvertedSalary'].fillna(
    round(df['ConvertedSalary'].mean())
)
Feature Engineering for Machine Learning in Python

一起來練習吧！

Feature Engineering for Machine Learning in Python

Preparing Video For Download...