Sample-Optimal PAC Learning of Halfspaces with Malicious Noise

                                 Jie Shen 1

              Abstract                   Generally speaking, a large body of existing works study the
    We study efficient PAC learning of homogeneous          problem of learning halfspaces under label noise. This in-
    halfspaces in Rd in the presence of malicious          cludes early works on random classification noise where the
    noise of Valiant (1985). This is a challenging  ...