Data Reduction

Data Reduction

In Machine Learning and Data Mining

• How to speed up computation for our model?

There are two approaches:

• reducing the number of rows
• reducing the number of columns

Rows

The main approach is to randomly select a subset of the dataset

• Sampling

Columns

Main approach: remove dependent variables

• Correlation coefficient for two Quantitative Variables
• Chi-square Test of Independence for two Categorical Variables
• One-Way ANOVA F-Test for quantitative vs categorical case

Other techniques:

• Principal Component Analysis and Singular Value Decomposition

See Also

• Bivariate Analysis

Sources

• Data Mining (UFRT)