Identification of firing patterns and synchronization control in neural networks via machine learning techniques

• Grant number: 24/07051-4
• Support Opportunities: Scholarships in Brazil - Post-Doctoral
• Start date: March 01, 2025
• End date: February 28, 2027
• Field of knowledge: Engineering - Electrical Engineering - Industrial Electronics, Electronic Systems and Controls
• Principal Investigator: Elbert Einstein Nehrer Macau
• Grantee: Adriane Reis Brugnago
• Host Institution: Instituto de Ciência e Tecnologia (ICT). Universidade Federal de São Paulo (UNIFESP). Campus São José dos Campos. São José dos Campos , SP, Brazil

Abstract

Neuronal synchronization is frequently studied in dynamic systems and in neuroscience, due to the relevance of its understanding for understanding the perfect functioning of the brain, especially in neurodegenerative pathologies that involve the motor part of the cortex. We will investigate the effects of delay time and synaptic plasticity in a neuronal network, whose behavior is described by the Hodgkin-Huxley and Huber-Braun models. Our premise is that the application of a delay time can significantly reduce the synchronized activity of the neuron network. To study synchronization control, it is important to understand the mechanisms that lead to the emergence of synchronized activity in the network. To do this, we will apply machine learning techniques to train a set of artificial neural networks in order to identify synchronized firings in advance. Furthermore, we will use artificial intelligence to induce and control network synchronization by adjusting synaptic weights through genetic algorithms. In the end, we hope to be able to control or induce synchronization in two distinct networks through the techniques developed in this research.