QBoost for regression problems: solving partial differential equations

A hybrid algorithm based on machine learning and quantum ensemble is proposed to find an approximate solution a partial differential equation with good precision favorable scaling in the required number of qubits. The classical component consists training several regressors (weak-learners), capable solving approximately using learning. adapting QBoost solve regression problems build learners. We have successfully applied our framework 1D Burgers’ viscosity, showing that method really improves solutions produced by weak-learners. also implemented D-Wave Systems, confirming performance compared simulated annealing exact solver methods.