0

I need to create a model, that finds when there will be a significant chance of occurring High peak in the signal. My data has N length of inputs and corresponding N length of outputs. For example,

Input is [12398.2, 15031.89, 17760.46, 20489.03, 23217.6, 25754.98,.....,12456.9,....24567.98] corresponding output is [0,0,0,0,0,1,....,0,....1].

0 represents No Peak and 1 represents Peak.

So, I need to create a time series model to predict when the peak will occur. I explore more about the model selection. I came across a few links which give me a slight idea about generating the model.

One efficient method is to come with window size and analyze its local mean to predict whether the peak will occur in a particular window frame. In this case using ARIMA model helps. But still, I hope that there will another easy approach to do this task.

Any other possible method I could try to predict the peak?

Finding Significant Peaks Evaluations of different method

Madhi
  • 101
  • 2
  • Welcome to CrossValidated. There is no "one-size fits all" answer here. It really is system and data dependent. What can you tell us about the "physics" that drive the numbers? – EngrStudent Dec 11 '18 at 13:15
  • 1
    Sorry for not being specific. I figured one solution for this problem. I will see either that solution helps. If not, I will edit the question again and explain more data dependent. Thanks! – Madhi Dec 11 '18 at 13:30

0 Answers0