"Proving Variable Independence in a Network Using Markov Blanket"

Click For Summary

Discussion Overview

The discussion revolves around the concept of proving the independence of a variable in a Bayesian network using its Markov blanket. Participants explore the definitions and implications of independence within the context of probabilistic models.

Discussion Character

  • Exploratory, Technical explanation, Conceptual clarification

Main Points Raised

  • One participant inquires about a simple proof for variable independence given its Markov blanket, emphasizing the desire for a memorable explanation rather than complex probability calculations.
  • Another participant seeks clarification on the definitions of "a variable" and "the other variables in the network," suggesting that "a variable" refers to a node in the Markov blanket and "the other variables" are nodes outside of it.
  • A participant defines "a variable" as a node in the Bayesian network along with its conditional probability table, and states that "independent" is used in the standard statistical sense.
  • One participant expresses interest in the outcome of the inquiry and indicates a willingness to engage further once more information is available.

Areas of Agreement / Disagreement

The discussion does not reach a consensus, as participants are still seeking clarity on definitions and proofs related to variable independence and Markov blankets.

Contextual Notes

Participants have not yet established a clear definition of independence in this context, and there are unresolved questions about the nature of the relationships between variables in the network.

0rthodontist
Science Advisor
Messages
1,229
Reaction score
0
Is there a simple (meaning, memorable and not just a lot of crunching through probability formulas) proof that a variable is independent of the other variables in the network, given its Markov blanket?
 
Technology news on Phys.org
Were you able to get this yet?

Is "a variable", node A in the Markov blanket? And "the other variables in the network" any node that does not belong to the blanket?

If so, then define independent?
 
--No, I have not yet found this out

--A variable is a node in the Bayesian network, together with its conditional probability table

--"Independent" is used in the standard statistical sense
 
If you get it, let me know. I'd be interested to hear. I'm sure I'll realize it in due time, and get back to you if you don't to me first.
 

Similar threads

MHB Convexivity issues in a new type of network - NLP
  • · Replies 1 ·
Replies
1
Views
3K
Neural Networks Question about the Hebbs and Delta Rule
  • · Replies 1 ·
Replies
1
Views
3K
Markov processes (Weight Suspended equally by n Cables)
  • · Replies 4 ·
Replies
4
Views
2K
JS Variable Hygiene: Preventing Global Var Access
  • · Replies 13 ·
Replies
13
Views
3K
Undergrad Variable Normalization for different variable ranges....
  • · Replies 3 ·
Replies
3
Views
2K
High School Choosing which variable is the dependent or independent variable
  • · Replies 16 ·
Replies
16
Views
2K
Use of AI (ML/DL) in Science
  • · Replies 6 ·
Replies
6
Views
4K
Undergrad Bell vs Kolmogorov: Unravelling Probability Theory Limits
  • · Replies 93 ·
4
Replies
93
Views
8K
Undergrad Understanding Stress and Strain: The Relationship Between Two Variables
  • · Replies 9 ·
Replies
9
Views
2K
Graduate Why Does the Standard Model Use Six Dimensions for Probability Calculations?
  • · Replies 21 ·
Replies
21
Views
3K