Synthetic Data Generation

Hello everybody.

I would like to ask your opinion on the subject of Synthetic Data Generation.

I'm planning on kicking off my algotrading journey for a while now, but each time I get discouraged by the lack of high-quality data.

I was thinking about creating a micro-service first which could either

• create synthetic data drawn from existing data-set distributions (maybe create random time-series from long-tail/normal distribution)
• create an Agent Based Modeling (ABM) service
• maybe create a GAN ??

Does anybody has any experience in any of these? Does it worth a shot?

​

Thank you, each answer is highly appreciated.