Session 3 – P-Metaheuristics: Genetic Algorithms & Ant Colony Optimization | LSO 2025 Workshop

This video is the third session of the online workshop “An Introduction to Metaheuristics”, conducted on 14 December 2025 (4:00–6:00 PM) as part of the Large Scale Optimization Workshop (LSO 2025).

The workshop is hosted by the Brij Disa Centre for Data Science and Artificial Intelligence, Indian Institute of Management Ahmedabad (IIM Ahmedabad). It is designed for students, faculty members, and industry professionals, and focuses on building a strong conceptual foundation for designing and implementing metaheuristic algorithms.

Across the three-day workshop, widely used metaheuristic techniques such as Local Search, Simulated Annealing, Tabu Search, and Genetic Algorithms are discussed, with an emphasis on conceptual understanding rather than implementation details.

Workshop slides and resources:
https://sites.google.com/view/shubham...

This Session (Session 3) introduces population-based metaheuristics, which operate on a set (population) of candidate solutions and use collective search mechanisms to balance exploration and exploitation of the search space. The session contrasts population-based methods with single-solution metaheuristics discussed in the earlier sessions.

Topics covered in this session include:

• Fundamentals of population-based metaheuristics
• Generation and replacement of populations
• Search memory and diversity in population-based search
• Evolutionary Algorithms: population, fitness, selection, reproduction, and replacement
• Genetic Algorithms: representation, selection, crossover, mutation, and elitism
• Swarm Intelligence concepts
• Ant Colony Optimization: pheromone trails, solution construction, evaporation, and reinforcement
• Application of Genetic Algorithms and ACO to combinatorial optimization problems

This session completes the conceptual framework of the workshop by covering two of the most widely used population-based metaheuristic approaches and highlighting their differences from single-solution methods.

Instructor:
Shubham Keshri
Ph.D. Scholar
Department of Management Sciences
Indian Institute of Technology Kanpur

Reference:
Talbi, E. G., Metaheuristics: From Design to Implementation, John Wiley & Sons, 2009.