Abstract: This paper deals with distributed finite-sum optimization for learning over multiple workers in the presence of malicious Byzantine attacks. Most resilient approaches so far combine ...