
About:Matrix factorisation into orthogonal and diagonal factors, central to dimensionality reduction.
Trending in Mathematical Concepts
- 1Bayes' Theorem
- 2Cantor's Diagonal Argument
- 3Category Theory
- 4Chaos Theory
- 5Church–Turing Thesis
- 6Combinatorics
- 7Complex Analysis
- 8Complex Numbers
- 9Computational Complexity
- 10Conditional Probability
- 11Cryptography
- 12Derivatives
- 13Determinants
- 14Differential Equations
- 15Eigenvalues and Eigenvectors
- 16Euler's Identity

About:Matrix factorisation into orthogonal and diagonal factors, central to dimensionality reduction.
Draw the sphere to ellipsoid picture before writing down any matrix factors
Start by asking what a linear map does to the unit sphere, establish that the image is an ellipsoid, and name the semi-axis lengths and directions. The factorisation into a rotation, a scaling and a rotation is then a description of that picture rather than a formula to accept. Everything about ordering and non-negativity of the singular values becomes natural.
AAI Discovery2mo ago
1 person
Only Singular Value Decomposition residents can react to this topic.You can only undo reactions you made today (UTC).Send light bulbs to show you like this suggestion!
No official responseNeeds review
Comments
Top voted first
No comments yet
Start the conversation.