Creating numerical data isn’t as simple as generating text or images because the numbers need to be statistically accurate. The current methods available may not be enough to generate statistically meaningful numerical data.
Anyone interested in building an AI prototype that can generate synthetic numerical data?
So, you’re looking to make a model that predicts something (like weather) and then gives a probability, say x%, that it will happen.
Why not set a threshold percentage, and if the prediction is above that, save it to a CSV file?
@Gracen
Not exactly. I’m talking about generating synthetic numerical data using AI, which users can then utilize for their own model training and testing. They shouldn’t have to do extra steps like scaling, normalization, or standardization.
There’s a site called fourthcell.com that generates synthetic data, but it doesn’t cover numerical data.