#Bayes' Theorem Argumentation

24 messages · Page 1 of 1 (latest)

vague escarpBOT
ocean ocean
#

uh, this is offtopic I think.

#

Others , please agree or disagree.

opal island
#

you can use bayes' for arguments like this or at least i've seen it done here

median salmon
#

well the content is offtopic but if you rewrite it as propositions then you can maybe arrive at something

opal island
median salmon
#

yeah i meant that if you rewrite them as actual logical dependencies

opal island
#

$$P(K \mid E) = \frac{P(E \mid K) \cdot P(K)}{P(E)}$$

rough inletBOT
#

Bodhisattva

median salmon
#

eh, ok, i don't really know how to proceed with that
tbh it needs to be formalized more, like what variables do you want to associate probability to, what are the relationships between the variables etc.

#

and assumptions that you are willing to make

#

bayes theorem basically gives you a way to invert conditional probabilities, but it won't help if you don't assume some actual values for probabilities or bound on them within the assumptions

opal island
median salmon
#

if you're just going for true/false statements then bayes's theorem might not be needed as it deals with probabilities, logic might be enough

#

though if you want to express something like "it's less probable that x than it is that ~x" then it might help

opal island
#

problem is deductive expressions can't do "X is more likely than Y"

median salmon
#

yeah exactly

#

and also cannot express things like "it's more probable than not"

#

so if you need that then rewrite it as probabilities

opal island
#

@ocean ocean

#

you can delete this post btw

indigo sun
#

.close