Are you a frequentist or a Bayesian?
Islam News @ NYT- Swedish Artist Says Sketch of Prophet Aimed to Mock All Religions: A man who prosecutors say was the target of a murder plot says his caricature of the Prophet Muhammad was intended to poke fun at all sacred symbols.
- Nigerians Recount Night of Their Bloody Revenge: This week dozens of herdsmen slaughtered hundreds of people in a brutal act of sectarian retribution in Jos.
- Views of ‘JihadJane’ Were Unknown to Neighbors: The ex-husband of the woman who called herself “JihadJane” said she “was just like anyone else.”
-
Would you like your own blog hosted at Talk Islam? Email the admin at "salaam at talk islam dot info" !
Talk Islam uses Wordpress MU and the Prologue theme by Automattic
DISCLAIMER: Please see the About Talk Islam page for fair use issues and our privacy policy.
DONATE: Talk Islam needs your support!
aziz 5:42 am on May 2, 2008 Permalink |
going by the definitions here – http://tinyurl.com/46lp6z
I’d say that if we have a-priori information, we might as well use it, but i am still more “comfortable” as a frequentist and would tend towards it wheneverthe aprioris were not available, or (subjectively speaking) trustworthy.
its better to select from tools in the toolbox than be dogmatic about it. You dont hear people asking the question, are you a Pearson’s or a Spearman’s correlator?
aasem 9:01 am on May 2, 2008 Permalink |
Being closely attached with the field of Estimation Theory and Stochastic Processes, I would argue that both approaches cover their own domains well.
As far as Communication or Adaptive Processing is concerned, one has to be a Bayesian while testing an hypothesis or making a decision. However a Bayesian formulation of problem requires the knowledge of a priori probabilities which are usually assigned by frequentist interpretive methods.
But that is only from a scientist’s POV
As far as philosophy is concerned I am a Bayesian.
aziz 9:54 am on May 3, 2008 Permalink |
techically, a frequentist would answer either “1″ or “0″, whereas a bayesian could only answer “P”
razib 9:25 pm on May 4, 2008 Permalink |
ultimately we are all bayesians, but proximately we are frequentists?
it’s like programming languages; use what works.