It’s challenging to convert higher dimensional data to lower dimensions or visualize the data with hundreds of attributes or even more. Too many attributes lead to overfitting of data, thus results in poor prediction.
Dimensionality reduction is the best approach to deal with such data. A large number of attributes in dataset leads may lead to overfitting of datasets. There are several techniques for dimensionality reduction. The three popular dimensionality reduction techniques to identify the set of significant features and reduce dimensions of the dataset are
1) Principle Component Analysis (PCA)
2) Linear Discriminant Analysis (LDA)
3) Kernel PCA (KPCA)
Data Science Enthusiast, Developer, Neophilia