A Non-Negative Matrix Factorization Framework for Privacy-Preserving and Federated Learning | IEEE Conference Publication | IEEE Xplore