Fair Dimensionality Reduction and Iterative Rounding for SDPs
Morgenstern, Jamie ; Samadi, Samira ; Singh, Mohit ; Tantipongpipat, Uthaipon ; Vempala, Santosh
arXiv, Tome 2019 (2019) no. 0, / Harvested from
Dimensionality reduction is a classical technique widely used for data analysis. One foundational instantiation is Principal Component Analysis (PCA), which minimizes the average reconstruction error. In this paper, we introduce the "multi-criteria dimensionality reduction" problem where we are given multiple objectives that need to be optimized simultaneously. As an application, our model captures several fairness criteria for dimensionality reduction such as the Fair-PCA problem introduced by Samadi, et. al. 2018 and the Nash Social Welfare (NSW) problem. In the Fair-PCA problem, the input data is divided into $k$ groups, and the goal is to find a single d-dimensional representation for all groups for which the maximum reconstruction error of any one group is minimized. In NSW the goal is to maximize the product of the individual variances of the groups achieved by the common low-dimensinal space. Our main result is an exact polynomial-time algorithm for the two-criterion dimensionality reduction problem when the two criteria are increasing concave functions. As an application of this result, we obtain a polynomial time algorithm for Fair-PCA for $k=2$ groups, resolving an open problem of Samadi, et. al. 2018, and a polynomial time algorithm for NSW objective for $k=2$ groups. We also give approximation algorithms for $k>2$. Our technical contribution in the above results is to prove new low-rank properties of extreme point solutions to semi-definite programs. We conclude with experiments indicating the effectiveness of algorithms based on extreme point solutions of semi-definite programs on several real-world datasets.
Publié le : 2019-02-28
Classification:  Computer Science - Discrete Mathematics,  Computer Science - Data Structures and Algorithms,  Computer Science - Machine Learning,  Mathematics - Optimization and Control,  90C22, 90C27, 90C29, 90C49, 68Q25,  F.2.0,  G.2.0
@article{1902.11281,
     author = {Morgenstern, Jamie and Samadi, Samira and Singh, Mohit and Tantipongpipat, Uthaipon and Vempala, Santosh},
     title = {Fair Dimensionality Reduction and Iterative Rounding for SDPs},
     journal = {arXiv},
     volume = {2019},
     number = {0},
     year = {2019},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1902.11281}
}
Morgenstern, Jamie; Samadi, Samira; Singh, Mohit; Tantipongpipat, Uthaipon; Vempala, Santosh. Fair Dimensionality Reduction and Iterative Rounding for SDPs. arXiv, Tome 2019 (2019) no. 0, . http://gdmltest.u-ga.fr/item/1902.11281/