I'm in the process of improving an algorithm. My improved version of the algorithm converges to the local optima with less number of iteration but it takes slightly longer in terms of convergence time in sec. Is it a contribution? How can defend it?
Dear Eahsun Fahim
This might happen due to large data and the high computational cost of formulations at each iteration. using Matlab software, ignoring loops might affect the process and increase the performance.
Sina
An important contribution--which your method just might be--is one that works best for a wide range of problems (few methods can achieve this) or works much better than the previous methods for an important class of problems. So try your method out on a bunch of problems and see what happens. Document each step so that when you have accumulated enough information you can publish the details. I have attached a paper on a nonlinear optimization algorithm that includes four very different test cases. I also wrote a book on the subject which is free today (4/13) https://read.amazon.com/kp/embed?asin=B09HHFDRZB&preview=newtab&linkCode=kpe&ref_=cm_sw_r_kb_dp_NTBG4ERXSG76MS9ZHJQF
- Badges
- Science topic
- Similar topics
- Human-Computer Interaction
- Collaboration