Review Note

Last Update: 01/08/2024 04:06 PM

Current Deck: Data Visualisation

Published

Currently Published Content

Front

What are the different forms of amputation?

Back

Row-wise deletion (<5% missing data)
Column-wise deletion (>40% missing data)
Imputation (in-between 5% and 40% missing data)
Pair-wise deletion (remove items only from analysis over missing values)

Problem: analysis over different variables may be based on different item sets

1 & 2 are rules of thumb. But not "strict"

You can still gain some results from remaining data only when -

Data is missing randomly
Dataset is large enough

Current Tags:

preprocessing week5

Pending Suggestions

No pending suggestions for this note.