Respected RG scholars, I have discovered that the computational complexity in nonlinear models available in many applications like financial planning, data analysis, risk management, scientific modeling, e.t.c. often requires global search approach. Please recommend the best....
Dear Modestus,
Truth is told in this question the best technique of global optimization depend of application field, here are some links with an real application; i think it could be inspired you to choose the best technique.
-Novel Heuristic Algorithm for Large-scale Complex Optimization
https://www.sciencedirect.com/science/article/pii/.../pdf?...
-The Future of Optimization Technology - University of Melbourne
https://people.eng.unimelb.edu.au/pstuckey/papers/position.pdf -Global Optimization Methods based on Tabu Search - OPUS Würzburg
https://opus.bibliothek.uni-wuerzburg.de/.../Global_optimization_methods_based_on_...
-Maple GOT - Semantic Scholar
https://pdfs.semanticscholar.org/.../d2dd916c57e4bf306f130232d2...
-adaptive random search methods for simulation optimization
https://pdfs.semanticscholar.org/8caf/61d95d77844fc3f888d7f2c57087add566c3.pdf
-Stochastic Optimization - The Johns Hopkins University Applied ...
www.jhuapl.edu/.../Comp_Stat_handbook_2nd-edition_Spall.pdf
Best regards
Dear Dr. Mohamed-Mourad,
Thank you for the valuable information you have provided. They are really helpful.
Regards
Modestus
Dr. Moussaab Belarbi,
Thank you for your inputs. Am glad to view all, the references are useful.
Regards
Modestus
Dr. Hamit Can,
Thank you so much for your time and valuable information. I appreciate...
Regards
Modestus
Hi,
It greatly depends on what can of problem you have to solve. If you assume that your problem is non convex, multimodal or ill conditionned, it is prefered to consider a algorithm like CMA-ES as it provide good invariance properties and is robust over a wide variety of problem classes, including multi-modal.
Whatever algorithm you decide to use: from CMA-ES to Gradient based approaches (like BFGS), it is recommended to to an appropriate benchmark to ensure to you will be able different type of problem closely related to industrial problems: for that, please look for COCO/BBOB black box optimization benchmark
Thank you very much Nacim Belkhir for your valuable information. There is really a lot to do with CMA-ES. I will contact you on further information on the algorithm.
Regards
Modestus
@ N. Belkheir, the problems is tied to transportation and transshipment. See the link : Article A comparative study of artificial neural network (ANN) and a...
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling