0

I have this idea for a prediction algorithm that predicts a random value with accuracy based a sequence of previous occurrences of an item from the set being chosen and analyse the pattern to improve the accuracy.

So basically an algorithm that accepts two parameters, one being a set of possible choices; the other being an history of those numbers, analyse that pattern and predict the next number in the sequence.

  1. Does this make sense to anyone else?
  2. If it does, what would be your approach to writing such an algorithm?
  3. What language would you refer?

The goal is pattern recognition in overlapping sample sets, find a correlation and establish an equation to predict that with some degree of accuracy.

Improve this question

1 Answer 1

Reset to default
1

Well, there are thousands of algorithms based on this approach. For example look at the formula for Shortest-job-first to predict the CPU burst of next process based on the previous Processes time and their burst because short-term-schedular has to take decisions fast.

Formula is:"T(n+1) = a*t(n) + (1 − a)Tn.

Were T(n+1) will be our predicted time, t(n) is the length of nth cpu burst. a is the relative weight of recent and path history, for example if a = 0 then next time is t(n).

For language it depends on you what language you are good with.

Improve this answer
Sign up to request clarification or add additional context in comments.

1 Comment

Thanks, Could you explain the concept behind it? I don't want to just copy a formula/algorithm. I want to understand how and why it works as well.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.