English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Tabular data synthesis has received wide attention in the literature. This isbecause available data is often limited, incomplete, or cannot be obtainedeasily, and data privacy is becoming increasingly important. In this work, wepresent a generalized GAN framework for tabular synthesis, which combines theadversarial training of GANs and the negative log-density regularization ofinvertible neural networks. The proposed framework can be used for twodistinctive objectives. First, we can further improve the synthesis quality, bydecreasing the negative log-density of real records in the process ofadversarial training. On the other hand, by increasing the negative log-densityof real records, realistic fake records can be synthesized in a way that theyare not too much close to real records and reduce the chance of potentialinformation leakage. We conduct experiments with real-world datasets forclassification, regression, and privacy attacks. In general, the proposedmethod demonstrates the best synthesis quality (in terms of task-orientedevaluation metrics, e.g., F1) when decreasing the negative log-density duringthe adversarial training. If increasing the negative log-density, ourexperimental results show that the distance between real and fake recordsincreases, enhancing robustness against privacy attacks.

Invertible Tabular GANs: Killing Two Birds with OneStone for Tabular Data Synthesis