Re: probability vs. likelihood

On Mon, 10 Aug 2009 10:12:30 -0700 (PDT), khacker
<kenneth.hacker@xxxxxxxxx> wrote:

I recently heard some government analysts argue that risk prediction
should be based more on likelihood than on probability. I never heard
this distinction before and wonder what primer I can read about the
differences.


Well, the simplest version of the distinction is the difference
between looking at the probability density - the height of the
curve - and looking at the CDF (cumulative density function).
Those *have* been applied somewhat differently when used
as a basis for inference, so I expect that you really want to see
something that talks about that.

I read the reply by Keo, and I don't object to anything in it
but it seems overly abstract.

Here is a link to a post from 2000 that discusses likelihood, etc. --
http://groups.google.com/group/sci.stat.math/msg/947946a120eef4d0?hl=en
and you might click to expand to see the rest of the thread.

I am no Bayesian, but I *think* that a Bayesian might say, "Use
likelihood more than probability", when advocating the use of
Bayesian analysis. If that was the context, then you may want
to see an intro to Bayesian statistics.

I surely don't know how this would apply, in particular, to any
variety of risk analysis. When I read "risk prediction", I thought
of "risk analysis", which I associate particularly with formal
analyses of safety of nuclear power plants, and of post-hoc
analyses of shuttle disasters. Under "risk prediction" more
generally, I also think of weather catastrophes and insurance
planning.

--
Rich Ulrich
.

Relevant Pages

  • Re: behavior as mapping
    ... experimentation favored by EABers is sort of consistent with Bayesian ... probability", posterior because it is the probability estimate AFTER ... P- the prior probability of the ith model in the distribution ... piecewise stable states (discrete jumps between some number of steady ...
    (comp.ai.philosophy)
  • Re: Gradual Learning, not Reinforcement Learning
    ... perhaps depending on a threshold probability level, ... That would be where the "utility model" comes in (moving from Bayesian ... book, the one on "Causality". ... that this is not "must" reading for behaviorists. ...
    (comp.ai.philosophy)
  • Re: Bayesian continued and shuffling
    ... From a Bayesian point of view did I not use prior knowledge? ... An observation about hand shuffling: ... For the record I try not to adjust and keep using pure probability ...
    (rec.gambling.poker)
  • Feasibility of bayesian network AIs
    ... I have recently begun reading about Bayesian belief networks. ... the conditional probability of the latter node given the former. ... "hard evidence" is assigned to the ...
    (rec.games.roguelike.development)
  • Re: when can bayesian "not" be applied?
    ... Bayesian learning and revision of Bayesian models. ... probability distributions, much of it about sequences. ... relative to the size of the "sequence space" then simple estimators (e.g. ... Now, sometimes all you really need is the entropy of a distribution, not the ...
    (comp.ai.philosophy)
Loading