The 8088 The 8088 ← All news
arXiv cs.LG AI Research Apr 23

Generative Augmentation of Imbalanced Flight Records for Flight Diversion Prediction: A Multi-objective Optimisation Framework

★★★★★ significance 2/5

This research proposes a multi-objective optimization framework to use generative models for augmenting imbalanced flight records. By creating synthetic diversion data, the study demonstrates improved predictive accuracy for rare aviation events using models like TVAE and CTGAN.

Why it matters Synthetic data generation offers a viable pathway to training reliable predictive models for high-stakes, low-frequency aviation safety events.
Read the original at arXiv cs.LG

Tags

#generative models #aviation #synthetic data #imbalanced data #optimization

Related coverage