Photonic reservoir computing improves performance with smaller output space
Photonic reservoir computing with dimensionally compressed readout
Neural and Evolutionary Computing
Summary
Photonic reservoir computers face a problem: their output parts can't be too big because of hardware limits. The authors studied a way to shrink the output data by mixing it randomly, keeping important information while fitting into a smaller space. They tested this idea by comparing a small system to a bigger one that compressed its output to the same size. Their results showed the smaller compressed system often works better, especially for certain tasks and settings. This approach helps make photonic reservoir computers more scalable by overcoming output size limits.
Reservoir computingPhotonic computingTime delay reservoirRandom projectionDimensionality reductionInformation processing capacityNARMA10 taskHardware constraintsHyperparameter tuningScalability
Authors
Gerald Kobi, Mohab Abdalla, Miguel C. Soriano, Damien Rontani
Abstract
This work addresses a hardware constraint in reservoir computing: the limited size of the readout layer imposed by systems with a physical readout. We investigate a strategy to accommodate this constraint based on random projection, which compresses high-dimensional reservoir states into a lower-dimensional subspace while preserving key properties of the source space and information- processing capabilities. To evaluate this approach, we compare a small, standalone time delay reservoir against a larger configuration whose output is projected down to match the same restricted readout dimension. Using task-independent metrics, we demonstrate that the distribution of information-processing capacities may differ between the two configurations, even at identical readout sizes. Furthermore, we perform a comprehensive hyperparameter scan to assess how both systems behave under varying physical regimes. Finally, we benchmark this approach on the standard NARMA10 task, showing that the random projection framework can yield superior performance compared to a standalone constrained reservoir, within specific compression range. These results provide a scalable pathway to bypass physical readout bottlenecks in hardware-based reservoir computing.