Multiclass Land Use / Land Cover (LULC) Classification Using Quantum Enhanced Support Vector Machines

Abstract

This work will present the challenges in using quantum-enhanced support vector machines (QSVM) for classification tasks on multi-spectral Earth Observation (EO) data. The main areas of investigation include the challenges in preparing and encoding the classical data into useful quantum states and training QSVMs at scale on gate-based quantum software simulators in HPC environment. Finally, some results comparing classical and quantum enhanced SVM models will be presented.