Estimating the Support of a High-Dimensional Distribution

John C. Platt, Bernhard Schölkopf, John Shawe-Taylor, Alex J. Smola, and Robert C. Williamson

November 1999

Suppose you are given some dataset drawn from an underlying probability distribution P and you want to estimate a "simple" subset S of input space such that the probability that a test point drawn from P lies outside of S is bounded by some a priori specified v between 0 and 1. We propose a method to approach this problem by trying to estimate a function f which is positive on S and negative on the complement. The functional form of f is given by a kernel expansion in terms of a potentially small subset of the training data; it is regularized by controlling the length of the weight vector in an associated feature space. The expansion coefficients are found by solving a quadratic programming problem, which we do by carrying out sequential optimization over pairs of input patterns. We also provide a preliminary theoretical analysis of the statistical performance of our algorithm. The algorithm is a natural extension of the support vector algorithm to the case of unlabelled data.

Publication type | TechReport |

Number | MSR-TR-99-87 |

Pages | 30 |

Institution | Microsoft Research |

- A Constructive RBF Network for Writer Adaptation
- Learning from the Wisdom of Crowds by Minimax Entropy
- Speaker Identification using a Microphone Array and a Joint HMM with Speech Spectrum and Angle of Arrival

> Publications > Estimating the Support of a High-Dimensional Distribution