Nov 8th, 2013, 3:00 pm in William Penny 212

Critical Spiking Neural Networks

Artificial intelligence is used as an umbrella term describing our attempts at reproducing two key human characteristics, intelligence and thought. We know that these abilities come from the brain, so why not start by reproducing the human brain ? A spiking neural network is an attempt at modelling how neural cells in nature behave dynamically. By reproducing the electrical signals of a single neuron and connecting neurons together we observe a complex interaction of spiking activity possibly underlying the intelligent behavior we see in nature. In my talk I will briefly describe this technique and give several examples on its application in neuroscience and artificial intelligence. I will conclude by defining what a critical network is and how chaotic behavior plays an important role in optimizing information processing in neural networks. This talk will be for those of you who are curious on how to make an artificial “brain” and how the chaos in your mind not only makes your forget where your house keys are but also optimizes your thought processes.

Filipe Peliz Pinto Teixeira is a PhD student in the Department of Computing at Imperial College London working on spiking neural networks and their applications within the fields of artificial intelligence and neuroscience. In his research he focuses on how chaos may optimize information flow and storage within neural systems. Filipe has a BSc in Computer Science from the University of Johannesburg in South Africa and an MSc in Artificial Intelligence from Imperial College London. In his spare time he can often be found running around with an over sized camera or attempting to reach a comatose state in front of his precious computer. More about me→