DCGAN to model spatial and temporal distributions of real ride sharing data
- 0 Collaborators
Developing a novel Generative Adversarial Network to synthesize data provided by Uber, Facebook and Emirates. ...learn more
Overview / Usage
The system is designed to take a wide variety of data primarily driven by ride sharing data collected from major cities. Using this data the temporal and spatial distributions will be captured and synthesized using a novel DCGAN structure. The data generation will be guided by a spectrum of conditional, meta-data inputs that will allow for generation of previously unseen scenarios modeled on a variety of given parameters.
The work is still in progress and I am unable to share further details on a public forum.