User Tools

Site Tools


public:grl_readingmemo

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision
Previous revision
public:grl_readingmemo [2024/04/04 15:10] – created liangpublic:grl_readingmemo [2024/04/04 16:12] (current) liang
Line 12: Line 12:
 ===== #01, liangz ===== ===== #01, liangz =====
  
-range: from the title page to p8 (end of Chapter 1)+range: from the title page to p8 (end of Chapter 1)
 + 
 +** NOTICE: ** please read both the scanned pictures and this memo.
  
 i: title page i: title page
Line 38: Line 40:
  
 p1: Chapter 1 Introduction p1: Chapter 1 Introduction
-  * node, edge, graph => ask the audience to illustrate some graphs +  * node, edge, relations, graph => ask the audience to illustrate some graphs 
-  * Zachary Karate Club Network (1977)+  * Zachary Karate Club Network (1977) and on its importance 
 + 
 +p2: 
 +  * It mentioned "a dramatic increase in the quantity and quality of graph data in the last 25 years." => Why 25 years? (hint: It means since 1995) 
 +  * ML is not the only way but may be interesting. 
 +  * adjacency matrix, adjacency list, simple graph and {0,1} 
 +  * (optional) graph processing from adjacency list to adjacency matrix seems an evidence showing human consumes matter and energy to increase its order. 
 + 
 +p3: 
 +  * multi-graph (variable number of types/relations) vs multi-relational graph (fixed number of types/relations) 
 +  * Heterogeneous graph (inner edges important than inter edges), multipartite graph (no inner edges, only inter edges), multiplex graph (inter edges important than inner edges) 
 +  * attribute or feature 
 + 
 +p4: 
 +  * graph (abstract structure) and network (real-world data) 
 +  * supervised (predict an output) and unsupervised (infer pattern) 
 +  * node classification: predict the label of a node given a small number of labelled nodes (|V_train| << |V|) 
 + 
 +p5: 
 +  * applications: bot detection in a social network, function of proteins in the interactome, classify the topic based on links, etc 
 +  * difference from a standard supervised learning: the assumption/bias of iid (independent and identically distributed) or no. 
 +  * popular inductive bias used in graph learning: homophily (same attrubute with neighbors), structural equivalence (similar local structure -> similar label), heterophily (e.g., gender). 
 + 
 +p6: 
 +  * supervised learning and semi-supervised learning, and GL (no iid assumption) 
 +  * relation prediction: e.g., recommendation system, side-effect. Notice the requirement of inductive bias. 
 + 
 +p7: 
 +  * clustering and community detection 
 +  * graph classification, regression, and clustering (to the audience: what is the general difference?
 + 
 +p8: 
 +  * iid assumption and why? -> Li-Yang 
 +  * Additional comment: Causal relation and correlation. ML is often consider the latter approach but actually we need to consider the former.
public/grl_readingmemo.1712211047.txt.gz · Last modified: 2024/04/04 15:10 by liang