Deep Learning Research
Linear Factor Model, Laten Variables
Types of LFMs
Using joint distributions for classification and
sampling
Description Given an undirected graph G = ( V, E ), a clique S is a subset
of V such that for any two elements u, v ∈ S, ( u, v ) ∈ E.
Using the notation ES to represent the subset of edges which have both
endpoints in clique S, the induced graph GS = ( S, ES ) is complete.
Finding the largest clique in a graph is an NP-hard problem, called the
maximum clique problem (MCP).
Cliques are intimately related to vertex covers and independent sets.
The concept of a latent variable
Additional learning of PCA
• https://slides.com/miteshkhapra2/cs6015_lecture20/fullscreen#/0/16/1
Acknowledgement
• https://www.cse.iitm.ac.in/~miteshk/CS7015/Slides/Teaching/pdf/Le
cture19.pdf
• https://github.com/mlberkeley/Deep-Learning-Decal-Fall-2017
• https://www.youtube.com/watch?v=spdjqjPD3sU