The multiple instance problem arises in tasks where the training
examples are ambiguous: a
single example object may have many alternative feature vectors
(instances) that describe it, 数据挖掘研究院
and yet only one of those feature vectors may be responsible for the
observed classification of
the object. This paper describes and compares three kinds of
algorithms that learn axis-parallel 数据挖掘研究院
rectangles to solve the multiple-instance problem. Algorithms that
ignore the multiple instance
problem perform very poorly. An algorithm that directly confronts the
multiple instance problem 数据挖掘研究院
(by attempting to identify which feature vectors are responsible for
the observed classifications)
performs best, giving 89% correct predictions on a musk-odor
prediction task. The paper also 数据挖掘研究院
illustrates the use of artificial data to debug and compare these
algorithms. 数据挖掘实验室
1 Introduction
Consider the following learning problem. Suppose there is a keyed
lock on the door to the supply
room in an office. Each staff member h... 数据挖掘研究院
