I'm using Kalman Filter approaches and I've just implemented the extended Kalman filter (EKF) with my object 2D trajectory. However, I have a mess of alternative approaches that may fit better like Unscented Kalman Filter (UFK), particle filters, adaptive filtering, etc.

How can I choose the most suitable algorithm for my case? In addition, are there algorithms that can predict more than one step ahead?

  • $\begingroup$ Hello. Welcome to AI SE! I've added a few links to the corresponding Wikipedia pages for more context about these concepts. Please, make sure that I've not changed anything. Feel free to add also a link to some reference that describes "Unscented Kalman Filter (UFK)", which I've not added. Moreover, maybe you should describe more in detail your problem. "2d trajectory prediction" is not very descriptive. Maybe if you provide more details about your specific problem the answer will be more specific. $\endgroup$ – nbro Jan 8 at 11:22
  • $\begingroup$ Thanks for the links. I'll change the title. My question is more advice about which approach to go with. $\endgroup$ – R2D2 Jan 8 at 11:47
  • $\begingroup$ All predictors can predict more than one step using recursion. $\endgroup$ – Brian O'Donnell Jan 8 at 14:12
  • $\begingroup$ As UKF it's recommended over EKF for efficiency. Which is the best implementation (python vs acc/cpu efficiency? $\endgroup$ – R2D2 Jan 8 at 17:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.