how can determine a best theshold value from a vector

I have vector of 8 element , let A=[0.5807 0.7627 0.6584 0.6274 0.6240 0.6195 0.6012 0.7334] , the corresponding func for each element of a is B=[ 1 0.7 0.8 1 1 1 1 1] .how can find the threshold value of a such that B=1 always. vector A is changed continouly so, i cannot take a specific value as a threshold. how can do it in matlab.

[Image Analyst]

No idea. How did you get the first B vector? What do you mean that the elements of B are a "func" for the elements of A? How are the different A created? If A is thresholded to create B then B would have only true or false values, not floating point values. If you want B to always = 1 then just do this

B = ones(1, length(A));

tutorial-how-to-ask-a-question-on-answers-and-get-a-fast-answer

[hanadi abbas]

sorry, my question may be not clear.

let B vector is an accuracy for each value of A vector. I want threshold value from vector A such that take the elements of A that maximize accuracy and ignor the others. B is not a function of A , so i canot use it as objective function. can I use PSO ?

I have parts of program, in each part there is new vector A . consequently, new accuracy vector B is obtained. so, in each part i want new threshold value that depend on accuracy.