Confidence in a rule

#1
For each record of data, I need to update a rule which is represented as a set of possible ranges.
For example if I get the following recored values: 1,2,3,10,11,12,13 my rule will be - [1,2],[10-13]
I create the possible ranges of data online, that means that every record of data may change the existing rule.
For example: I get the following record values one after the other: 1,2,3,2,10:
First the ranget will be [1], then [1,2], then [1,3] then still [1,3] and then ([1,3],[10])

In each step I want to calculate my confidence in the ranges I collected.
This of course depends on the number of occurrences that have not changed the existing range. For example the confidence in the range [1,3] is greater than the range [10] , in last example, since value 2 didn't cange the range [1,3], but only strengthened the fact that the range is true.

For example, see the following example: 1,2,3,4,2,3,2,3,10
the result is the set of ranges is [1,4],[10]. Of course, the confidence in the range [1.4] is greater than the confidence in the range [10].

Do you have any idea on how can I calculate the confidence?
Thank you.
 
Last edited:

hlsmith

Not a robit
#2
@missRob - I could not fully understand your posted question. If you are not proficient in English, perhaps you can get someone to help you write out your questions. The above text has many punctuation and typographical errors.

Thank you.
 
#3
@missRob - I could not fully understand your posted question. If you are not proficient in English, perhaps you can get someone to help you write out your questions. The above text has many punctuation and typographical errors.

Thank you.
Sorry, I had problem with my PC and it mix the words and combine old ones with new. Hope you can understand now.
 

Dason

Ambassador to the humans
#4
For a little bit I thought maybe you were just trying to estimate what values are actually in your population but you say things like
For example the confidence in the range [1,3] is greater than the range [10]
which doesn't really make sense in that context.

Can you elaborate on 1) How you're defining "range" 2) What your overall goal is 3) if this is an abstraction of some problem... tell us what the actual problem is because I'm not convinced you've translated it well if that's the case.