Data mining (data preprocessing) -- Notes

data mining

Data cleaning

Anomaly detection

1. Missing value fill

local Outlier Factor
distancek(A,B)=max(diatancek(B), d(A,B))
K a near neighbor .
You can't directly compare Lrd(A) Size
To do ratio LOF（A）

The following data is the displayed result ：
The bigger the data , The more outlying .

2、Duplicate Data ( Duplicate data ) To weed out

Create Keys -> Sort -> Merge
keys key word -> Used to find two relatively close data .

Data Transformation