- This topic has 3 replies, 3 voices, and was last updated 1 year, 1 month ago by
.
- Posts
Well thanks for sharing that…boy is this a crazy market! Yeah, the real value I’ve seen in this over the years is missing out on this big drops.
I still think our BTC and ETH models could have handled the start of this better. That said, it’s hard to train on a scenario such as the Coronavirus.
Also, we should have the news on our 5.x model out today or tomorrow. It has been running in alpha for the last 30 days and has handled this “edge case” scenario very well.
I’m excited to get it out.
The AI may not have been able to avoid the pandemic situation completely, but I would rather lose a small % of my money from incorrect reversal predictions with a tight stop loss, rather than buying at near 8k levels and then getting rekt with this sharp drop. Besides, the AI will learn about the parameters that are reminiscent of “edge cases” such as this black swan event, and not make the same mistake in the future should such events happen again.
But then again, such black swan events will generally rekt most traders and investors because TA charts go flying out of the window; you simply cannot win by TA alone, since its based on fundamental factors like extreme panic selling for survival and human irrationality.
The timing on this thread is interesting. Rob D posted it after the first wave (which is now minor in comparison). Then we got hit with the tsunami.
@glennseer, I totally agree. I just know machine learning can do better. In fact, our 5.0 models have been deployed in an alpha state (running but not issuing trades). This is to ensure the entire operation is reliable before we hook them up to the customer notification system.We got to see how they performed over the last 24-48 hours. The Ethereum actually made it out with a 12% gain. We will have it live within a day. Bitcoin took a minor loss, but also handled it very well, all things considered.
You probably just received an email on it, but for detail: https://crypto-ml.com/blog/machine-learning-upgrade-to-5-0-deep-neural-networks/
- You must be logged in to reply to this topic.