By: Amir Rahamni, Instytut Fizyki PAN, Warsaw, Poland
When: December, 3rd, 12PM
Where: Sala de Juntas, ICMM
Abstract: Quantum reservoir computing (QRC) has emerged as a powerful paradigm for tackling machine-learning tasks by using the rich dynamics of quantum systems. At its core, QRC relies on effective control drives to encode input data and nonlinear mappings to transform these inputs into output features. Photonic platforms are likely candidates for QRC. However, they often lack sufficient nonlinearity for optimal performance. To overcome this limitation, a hybrid optoelectronic approach can be used, combining the speed of optical processing with the inherent nonlinearity of electronics. In this seminar, we present a method to enhance nonlinearity by employing wavefunction engineering within the regime of light-matter coupling. We explore various structures, including GaAs and TMD materials in cavities and waveguides. Through some examples, we demonstrate improvements in the accuracy of some quantum tasks, highlighting the potential of QRC in advancing machine learning.
