I’m planning to dig more into deep learning and while browsing google results I discovered these two blog posts quite interesting:
From the very quick overview looks like people are training models based on daily information about exchanges (open, high, low, volume) and of course my plan here is to try train a model on all of the (recent) trades that took place per each pair. Maybe there’s a deeper reason on why to train on more generic data than granular one, but it’s still ahead of me to find out.
I don’t think 1 hour will be enough to familiarize with great information gathered in above posts and moreover a friend of mine just let me know there’s promotion for a game and I’d like to at least install and try it for a few minutes. Distractions..
Thanks for reading,