0

I hope that my question is clear enough to understand.

If we need to calculate the probability of an event and we have historical data regarding our parameter of interest - When we have to assume distribution in order to calculate the probability of an event, or it is safe just to calculate the probability from a simple frequency distribution.

To put the question into context here is an example:

We want to calculate the probability of X numbers of sales in a certain period of time. Assume Poisson distribution? Or it is enough and "safer" just to calculate the probability of X sales from the historical frequency of sales during that period?

Thank you very much!

Lex
  • 61
  • 2
  • The answer is that your prediction method will depend on a lot. What are you trying to do? – Dave Mar 01 '20 at 23:12
  • Most probability calculations depend on parameters of a distribution (binomial, normal, Poisson, etc). Historical data provides a basis for parameter estimation, which are then used to make predictions (probability estimates for a new set of parameters). An example is the binomial distribution: if 13 people out of 72 people got sick last year from food poisoning at the annual picnic, how many might get sick this year if 100 people attend? –  Mar 01 '20 at 23:35
  • I marked it as a duplicate of another question. Please see if it answers your question, if not, please edit and tell us why? – Tim Mar 02 '20 at 06:31

0 Answers0