Image Clustering Using Linear Discriminant Analysis (LDA) Compared to t-SNE / UMAP

1. Is UMAP better than t-SNE?
2. Why is LDA better?
3. Is UMAP better than PCA?
4. What is the main difference between LDA and PCA?

Is UMAP better than t-SNE?

Thanks to the solution in building the high dimensional graph, UMAP theoretically saves more time and computation cost than t-SNE. It is reported that dimensionality reduction of a dataset from 784-D to 3-D took UMAP only 3 minutes, while it took t-SNE 45!

Why is LDA better?

The advantage of LDA is that it uses information from both the features to create a new axis which in turn minimizes the variance and maximizes the class distance of the two variables.

Is UMAP better than PCA?

UMAP outperformed t-SNE and PCA, if we look at the 2d and 3d plot, we can see mini-clusters that are being separated well. It is very effective for visualizing clusters or groups of data points and their relative proximities.

What is the main difference between LDA and PCA?

PCA is an unsupervised learning algorithm while LDA is a supervised learning algorithm. This means that PCA finds directions of maximum variance regardless of class labels while LDA finds directions of maximum class separability.