Private measurement of nonlinear correlations between data hosted across multiple parties

We introduce a differentially private method to measure nonlinear\ncorrelations between sensitive data hosted across two entities. We provide\nutility guarantees of our private estimator. Ours is the first such private\nestimator of nonlinear correlations, to the best of our knowledge within a\nmulti-party setup. The important measure of nonlinear correlation we consider\nis distance correlation. This work has direct applications to private feature\nscreening, private independence testing, private k-sample tests, private\nmulti-party causal inference and private data synthesis in addition to\nexploratory data analysis. Code access: A link to publicly access the code is\nprovided in the supplementary file.\n