Papers

20

Total Citations

214

H-Index

9

About

Shingo Mabu is a computational intelligence researcher whose work spans evolutionary algorithms, graph-based optimization, and intelligent robotics systems. His most significant contributions lie in the development and refinement of Genetic Network Programming (GNP), a graph-structured evolutionary framework that enhances traditional genetic algorithms through richer representational capacity. By integrating reinforcement learning into GNP, Mabu extended the paradigm's applicability to dynamic, real-world tasks such as mobile robot navigation and adaptive behavior generation in changing environments — work that has accumulated over 22 citations in its foundational form. Among his most impactful contributions is a graph-based Estimation of Distribution Algorithm (EDA) that leverages reinforcement learning to build probabilistic models, earning 37 citations and representing a notable advance in avoiding building-block disruption common in conventional evolutionary approaches. He further innovated by incorporating infeasible individuals into probabilistic model building, challenging the prevailing assumption that only high-quality solutions inform useful distributions. Beyond optimization theory, Mabu has made applied contributions in computer vision and robotics, including a deep learning-based citrus-harvesting visual system (30 citations) and gesture recognition using self-organizing maps. Together, his body of work — spanning over a decade — reflects a sustained effort to bridge theoretical evolutionary computation with practical intelligent systems applications.

Research Focus

Key Achievements

9
H-Index
20
Papers
214
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: 25
🏛 Institutions: Waseda University, Yamaguchi University, Advanced Research Institute

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 · 13 days ago