Abstract:
|
Modern data sets grow both in size and complexity. For example The Cancer Genome Atlas (TCGA) contains a large number of heterogenous data sets obtained by measuring different genomic phenomena, such as gene expression, copy number, snips, etc., on the same tissue samples. Integration of information across these disparate data sources together with uncertainty quantification is of great interest. In this talk we describe our ongoing work in using generalized fiducial inference to quantify the estimation uncertainty for model parameters of interest.
|