PCA Visualizer
Principal Component Analysis for dimensionality reduction
Configuration
Original Dimensions:
3
Number of Samples:
Components to Display:
2
Feature Correlation:
High Correlation
Medium Correlation
Low Correlation
Run PCA
PCA Results
Download Data
Principal Component Space
Scree Plot
Variance Explained
Component Loadings
Biplot (Observations & Variables)
Help
Principal Component Analysis
Goal:
Find orthogonal directions of maximum variance in data
PC1:
Direction with highest variance
PC2:
Direction with second highest variance, orthogonal to PC1
Dimensionality Reduction:
Keep top components, discard others
Understanding Results
Scree Plot:
Shows variance explained by each component
Loadings:
Contribution of original features to each PC
Biplot:
Shows data points and feature vectors together
Cumulative Variance:
Total variance captured by selected components
Elbow Method:
Look for bend in scree plot to choose number of components