“…reinforcement learning [3,6,10,52,73,81,88,114,138,172,191,204], and unsupervised clustering [17,18,41,89,140] or anomaly detection [43,140] algorithms. Furthermore, notable examples of poisoning attacks and defenses have also been shown in computer-security applications dealing with ML, including spam filtering [13,46,57,125,132], network traffic analysis [140], and malware detection [133,146,161], audio [1,90,106,109,192] and video analysis [166,208],…”