Machine Learning — Hard
Key points
- UMAP minimizes cross-entropy between high-dimensional and low-dimensional fuzzy set representations
- UMAP constructs a high-dimensional fuzzy topological graph from the data
- Optimization is done using stochastic gradient descent
- The objective is to find low-dimensional embeddings that best represent the data
- UMAP does not assume linear relationships like PCA
Ready to go further?
Related questions