Approximate NORTA simulations for virtual sample generation

作者：

Highlights：

• We introduce a sampling method that aims at generating general structured data.

• It accommodates for user-specified marginals and correlation structures.

• Both nominal and ordinal variables are handled.

• A Newton scheme ensures convergence to the desired correlation matrices.

• We create virtual samples so as to expand a small dataset used by a Fintech company.

摘要

•We introduce a sampling method that aims at generating general structured data.•It accommodates for user-specified marginals and correlation structures.•Both nominal and ordinal variables are handled.•A Newton scheme ensures convergence to the desired correlation matrices.•We create virtual samples so as to expand a small dataset used by a Fintech company.

论文关键词：NORTA simulation,Multivariate sampling,Regression trees,Support Vector Machine

论文评审过程：Received 31 August 2016, Revised 25 November 2016, Accepted 19 December 2016, Available online 21 December 2016, Version of Record 27 December 2016.

论文官网地址：https://doi.org/10.1016/j.eswa.2016.12.027