3 ms·
You can train a small NN on a bunch of synthetic data and the inference will be pretty fast.
by logicchains 3y ago
You can train a small NN on a bunch of synthetic data and the inference will be pretty fast.
- gituliar 3y agoCurious to see some references to people doing this. Have any ?
- logicchains 3y agohttps://www.aimspress.com/aimspress-data/qfe/2023/2/PDF/QFE-07-02-011.pdf https://www.aimspress.com/aimspress-data/qfe/2023/2/PDF/QFE-... is an example
- gully_boy 3y agowas thinking exactly this... and even better, why not on real data, especially since real, clean, data in this field is not at all hard to find.
- gituliar 3y agoReal historical data might not cover current market state. For example, option price among other params depends on the interest rate. For the last decade interest rate was around 0% in Europe and slightly higher in US. If you train on that data only, there is no chance to "learn" option prices in the high-interest-rate environment which we saw for the last few years. Hence, you need synthetic data to learn that region of the market space.