Related Experiment Video
Updated: Mar 30, 2026

Applying an eMASS Customization Program as a Research Tool to Evaluate Consumer Benefits
Published on: September 27, 2019
A global optimization paradigm based on change of measures.
Saikat Sarkar1, Debasish Roy1, Ram Mohan Vasu2
1Computational Mechanics Laboratory, Department of Civil Engineering , Indian Institute of Science , Bangalore 560012, India.
A new global optimization framework, COMBEO (Change Of Measure Based Evolutionary Optimization), uses derivative-free terms to enhance search accuracy and speed. This evolutionary optimization method offers a more rational and efficient alternative to existing schemes.
Area of Science:
- Computational Mathematics
- Optimization Algorithms
- Evolutionary Computation
Background:
- Global optimization is crucial for complex problems.
- Existing methods like particle swarm and differential evolution have limitations.
- Derivative-free optimization is needed for certain problem classes.
Purpose of the Study:
- To propose a novel global optimization framework named COMBEO (Change Of Measure Based Evolutionary Optimization).
- To introduce derivative-free additive directional terms using a change of measures.
- To integrate concepts from other global search methods into a unified framework.
Main Methods:
- Development of COMBEO framework.
- Incorporation of derivative-free additive directional terms via change of measures.
- Integration of particle swarm and differential evolution concepts.
- Application of random perturbations like 'scrambling' and 'selection'.
Main Results:
- COMBEO demonstrates a more rational and accurate approach to global optimization.
- Numerical results show COMBEO can be faster than existing evolutionary optimization schemes.
- The framework allows for adjustable search behavior (greedy vs. exploratory).
Conclusions:
- COMBEO offers a versatile and effective alternative for global optimization.
- The change of measures technique provides a robust foundation for the framework.
- The method's adaptability makes it suitable for diverse optimization challenges.
Related Concept Videos
Methods of Medium Optimization
Measurement: Derived Units
Area Computation by the Alternative Coordinate Method
Optimization Problems
Measurement: Standard Units
Scale-Up Processes
