Addressing Bias and Subjectivity in Machine Learning

Zhao, Yijun.
2017

Abstract: The success of supervised machine learning algorithms rests on the assumption that data are drawn from the same underlying distribution. However, this assumption is often violated in real world applications where collected data involves human judgement. The contribution of this thesis is a collection of approaches that address bias and subjectivity in real world data. We illustrate our w... read more

Subjects
Tufts University. Department of Computer Science.
Permanent URL
http://hdl.handle.net/10427/012562
ID: tufts:22476
To Cite: DCA Citation Guide