Regularisation and the inverse problem