/explore
|
You are here 
|
hadrienj.github.io | ||
| | | | |
austinmorlan.com
|
|
| | | | | It took me longer than necessary to understand how a rotation transform matrix rotates a vector through three-dimensional space. Not because it's a difficult concept but because it is often poorly explained in textbooks. Even the most explanatory book might derive the matrix for a rotation around one axis (e.g., x) but then present the other two matrices without showing their derivation. I'll explain my own understanding of their derivation in hopes that it will enlighten others that didn't catch on right away. | |
| | | | |
thenumb.at
|
|
| | | | | [AI summary] This text provides an in-depth exploration of how functions can be treated as vectors, particularly in the context of signal and geometry processing. It discusses the representation of functions as infinite-dimensional vectors, the use of Fourier transforms in various domains (such as 1D, spherical, and mesh-based), and the application of linear algebra to functions for tasks like compression and smoothing. The text also touches on the mathematical foundations of these concepts, including the Laplace operator, eigenfunctions, and orthonormal bases. It concludes with a list of further reading topics and acknowledges the contributions of reviewers. | |
| | | | |
blog.georgeshakan.com
|
|
| | | | | Principal Component Analysis (PCA) is a popular technique in machine learning for dimension reduction. It can be derived from Singular Value Decomposition (SVD) which we will discuss in this post. We will cover the math, an example in python, and finally some intuition. The Math SVD asserts that any $latex m \times d$ matrix $latex... | |
| | | | |
yang-song.net
|
|
| | | This blog post focuses on a promising new direction for generative modeling. We can learn score functions (gradients of log probability density functions) on a large number of noise-perturbed data distributions, then generate samples with Langevin-type sampling. The resulting generative models, often called score-based generative models, has several important advantages over existing model families: GAN-level sample quality without adversarial training, flexible model architectures, exact log-likelihood ... | ||
