In-materio reservoir working at low frequencies in a Ag<sub>2</sub>S-island network
Motoharu Nakajima, Kazuki Minegishi, Yosuke Shimizu, Yuki Usami, Hirofumi Tanaka, Tsuyoshi Hasegawa
Abstract
A reservoir that is more sensitive to lower frequencies is developed by a Ag 2 S-island network, where Ag filament growth/shrinkage achieves non-linear transformation of input signals. Six logic operations are achieved with accuracy higher than 99%.
Topics & Concepts
ShrinkageReservoir computingConductanceAmplitudeElectrodeMaterials scienceGeologyComputer scienceComposite materialPhysicsCondensed matter physicsOpticsArtificial neural networkArtificial intelligenceQuantum mechanicsRecurrent neural networkAdvanced Memory and Neural ComputingNeural Networks and Reservoir ComputingQuantum-Dot Cellular Automata