This is Noteaureus
Noteaureus.org is a place for notes about everything and anything that pops into jszym's head. At the moment, it's mostly a place to organise notes on deep learning papers, with a particular focus on graph learning.
The best way to browse the site is to take a look at the tags:
Notice any spelling mistakes or other inaccuracies? Shoot me a line at noteaureus [ΑΤ] jszym.com
Functional Discovery via a Compendium of Expression Profiles
Functional similarity based on co-expression
Relational Inductive Biases, Deep Learning, and Graph Networks
A pleasant review/positional paper that takes its time strolling through deep graph networks and how viewing them through the lens of relational inductive biases can be helpful.
Position-Aware Graph Neural Networks
A paper that addresses the problem of learning embeddings that encode for node position relative to the graph.
Setting Up a Windows System to Protect Untechnical Users from Themselves
Want to protect a friend or family from their untechnical selves? Here’s how I lock down systems for users with modest needs.
Multiple Sequence Aligning with STAR
Do you have a bunch of reads from an RNA-Seq experiment? Because I do.
Semi-Supervised Classification with Graph Convolutional Networks
Kipf, Thomas and M. Welling. “Semi-Supervised Classification with Graph Convolutional Networks.” ArXiv abs/1609.02907 (2017): n. pag.
On Spectral Clustering: Analysis and Algorithm
The paper which made spectral clustering practical. Useful for understanding other spectral techniques. Paper by M.I. Jordan & A. Ng., 2001