PCA is a dimensional reductionality technique used to handle data with high sparsity (data cursed with high dimensionality). The idea is to extract a limited number of significant features for training the model based on the extent of correlation between the features. The current repository involves in extracting 24 most significant vector components from a high dimensional data, visualizing them and reconstructing the images back again using only the signficant vector features to see the extent of information loss.
-
Notifications
You must be signed in to change notification settings - Fork 0
Unsupervised learning
License
vishnu-ek/Principal-Component-Analysis
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
About
Unsupervised learning
Topics
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published