Active Learning Under Malicious Mislabeling and Poisoning Attacks | IEEE Conference Publication | IEEE Xplore