Related Experiment Video
Updated: Apr 25, 2026
11:18
Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks
Published on: March 2, 2015
11.5K
Operant conditioning: a minimal components requirement in artificial spiking neurons designed for bio-inspired
André Cyr1, Mounir Boukadoum1, Frédéric Thériault1
1Computer Science Department, Cognitive and Computer science, Université du Québec à Montréal Montréal, QC, Canada.
Frontiers in Neurorobotics
|August 15, 2014
Summary
This study shows a simple artificial neural network can learn like biological agents using operant conditioning (OC). This bio-inspired approach simplifies robot learning in unknown environments.
Area of Science:
- Neuroscience
- Robotics
- Artificial Intelligence
Background:
- Operant conditioning (OC) drives behavioral changes in biological agents through reward and punishment.
- Autonomous robots in unknown environments can benefit from OC-like learning.
- Existing computational models for OC are often complex.
Purpose of the Study:
- To investigate operant conditioning (OC) learning in a bio-inspired paradigm using artificial spiking neural networks (ASNN).
- To demonstrate a simple neural micro-circuit can sustain OC in various learning scenarios.
- To propose a simpler alternative to existing computational OC models.
Main Methods:
- Utilized artificial spiking neural networks (ASNN) as robot brain controllers.
- Developed a minimal neural circuit integrating habituation and spike-timing dependent plasticity.
- Tested the OC module in multiple tasks of incremental complexity.
Main Results:
- A simple invariant micro-circuit was sufficient to sustain OC learning.
- The proposed OC module enabled ASNNs to perform various OC procedures.
- Minimal neural components achieved complex behaviors in simulated robot tasks.
Conclusions:
- A simplified bio-inspired OC module can be effectively implemented in ASNNs.
- This approach leads to simpler neural architectures for complex robot behaviors.
- The study offers a more accessible method for designing learning tasks in neurorobotics.
