While this is a much pursued field of research because of the promise of space and power-efficiency in neuromorphic computation the Southampton team claims a number of firsts with its work reported in Nature Communications .
In particular the Southampton team experimentally demonstrated an artificial neural network that uses memristor synapses supporting sophisticated learning rules in order to carry out reversible unsupervised learning of noisy input data. The ability to perform unsupervised learning in the presence of corrupted input data and probabilistic neurons, could be key to the development of robust big-data neuromorphic processors.
So-called memristors are electrical components that limit or regulate the flow of electrical current in a circuit and can remember the amount of charge that was flowing through it and retain the data, even when the power is turned off. These are essentially and physically equivalent to resistive RAMs, although resistive RAMs are more usually configured to store 1s and 0s.
Acting like synapses in the brain, the metal-oxide memristor array was capable of learning and re-learning input patterns in an unsupervised manner within a probabilistic winner-take-all (WTA) network. This is useful for enabling low-power embedded processors needed for the Internet of Things that can process in real-time big data without any prior knowledge of the data, the research team claimed.
Related links and articles: