As a follow-up to my previous Christmas blog post, Twas the night before big data, I give you ...
Bayesian network, the directed acyclic graphical (DAG) model
( yeah, yeah, I know it doesn’t quite have the same ring as last year’s title, but you'll recognize the tune)
You know Linear and Logistic Regression and Decision Tree and Neural Network,
You know TwoStage and MBR and SVM and Partial Least Squares,
But do you recall?
The most famous predictive model of all?
Has a directed acyclic graphic model with nodes representing random variables
And if you ever saw it
You would even say the links between nodes represent conditional dependency of the random variables
All of the other models used to laugh and call him names
They never let poor Bayesian networks
Join in solving any decision games
Then one foggy Christmas Eve,
The data scientist came to a say,
Bayesian networks with your different types of Bayesian network structures
Won’t you solve this problem with random variables under uncertainty?
Then all the models loved him,
As they shouted out with glee,
You’ll go down in history!
All kidding aside, follow these links to learn more about Bayesian analysis in SAS or SAS High Performance Bayesian networks, which is part of SAS High Performance Data Mining.
Happy New Year!