Machine learning in trading: theory, models, practice and algo-trading - page 3487
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
This is the harsh reality. The bare facts. The hopelessness of being. The plunge into inevitability.
Looks like my suspicions are correct
Yes, I have a link in my feed to a video course for beginners on coursera for free that explains it all. It's exactly entry level.
That's what I'm saying.
Looking at a MAE and thinking it's the equivalent of a profit is a beginner's trait.
Well you have stats on a set of models. The average stat should be at least better than the average stat at random, I wrote it. That's not bad enough for a cursory evaluation to see how randomly you get good forwards.
So I've shown the results of, say, two random samples, and on one of them the training is significantly better than on the other goes!
That's why it's not clear which random to base the results off of - that's the message.
That's impossible. It is quite possible to win on SB, and this, such luck, can last for a very long time. Take a test, see for yourself.
You have to make it possible...
So I showed the results of, let's say, two random samples, and on one of them the learning is significantly better than on the other goes!
That's why it's not clear which random sample to use - that's the point.
On average.
Then it is necessary to generate N random samples with different statistical distributions and settings at the beginning, to train a hundred models on each...
Then it is necessary to generate N random samples with different statistical distributions and settings at the beginning, train a hundred models on each one...
Here we need to make it possible....
Something like that
Too costly, there must be another way.