Kotaro Hirasawa

Waseda University, Kyushu University

Papers

16

Total Citations

178

H-Index

8

About

Kotaro Hirasawa is a computational intelligence researcher whose work sits at the intersection of evolutionary computation, estimation of distribution algorithms (EDAs), and reinforcement learning. His most significant contributions center on the development and refinement of Genetic Network Programming (GNP), a graph-based evolutionary framework that represents solutions as directed network structures rather than conventional string-based chromosomes, offering enhanced expressive power for complex optimization tasks. Hirasawa has made particularly notable advances in probabilistic model building within evolutionary algorithms, pioneering methods that leverage both high-performing and low-performing individuals to guide population evolution — a counterintuitive but demonstrably effective strategy explored across several of his studies. His 2013 paper introducing a graph-based EDA extended with reinforcement learning stands as his most influential work, accumulating 37 citations, with further contributions addressing critical challenges such as premature convergence, population diversity maintenance, and adaptability in dynamic environments. His applied research extends these frameworks to practical robotics, including mobile robot navigation using fuzzy logic and multi-stage reinforcement learning. Collectively, his body of work — spanning over a decade and garnering citations across the evolutionary computation community — has meaningfully advanced the theoretical foundations and real-world applicability of graph-based intelligent optimization systems.

Research Focus

Key Achievements

8
H-Index
16
Papers
178
Total Citations
11
Avg Citations/Paper
🏆 Most Cited Paper
A Novel Graph-Based Estimation of the Distribution Algorithm and its Extension Using Reinforcement Learning
37 citations · 2013
📈 Most Prolific Year: 2011 (4 Papers)
🤝 Key Collaborators: 17
🏛 Institutions: Waseda University, Kyushu University

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 15 days ago