Papers
20
Total Citations
251
H-Index
9
About
Milton Roberto Heinen is a Brazilian researcher whose work spans machine learning, neural networks, and autonomous robotics — fields he has approached with a consistent drive toward adaptive, biologically inspired computation. He is perhaps best known for his development of incremental learning models, most notably the Incremental Gaussian Mixture Network (IGMN), a neural architecture capable of continuously learning probability distributions from streaming data without requiring retraining from scratch. His 2010 paper on incremental learning of multivariate Gaussian mixture models has garnered 66 citations, making it his most influential contribution and a meaningful reference in the online learning community. Heinen also developed the Incremental Probabilistic Neural Network (IPNN), extending these ideas to regression and reinforcement learning tasks. Alongside this probabilistic learning thread, he pursued a parallel research agenda in legged robotics, designing the LegGen system, which uses genetic algorithms and neural networks — including Elman recurrent networks — to automatically generate stable gaits for simulated robots. His work on visual attention models further demonstrates his interest in robot perception. Together, these contributions reflect a researcher dedicated to building machines that learn and adapt in real time, with over 190 cumulative citations across his most recognized publications.
Research Focus
Key Achievements
Top Papers
- 1Incremental Learning of Multivariate Gaussian Mixture Models66 citations · 2010
- 2
- 3
- 4
- 5Neural Networks Applied to Gait Control of Physically Based Simulated Robots14 citations · 2006
- 6
- 7Evaluation of visual attention models under 2D similarity transformations10 citations · 2009
- 8
- 9Concept Formation Using Incremental Gaussian Mixture Models9 citations · 2010
- 10Applying Genetic Algorithms to Control Gait of Simulated Robots9 citations · 2007