Machine learning in trading: theory, models, practice and algo-trading - page 820
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
While there is a lull, I'll post some text here, maybe someone will be interested.
What does this have to do with the Ministry of Defense? There is also a topic "Interesting and Humor" - you can go there. Maybe someone will be interested.)
Thank you, very curious.
"§1.8 Conclusions
Everyone without exception... seeks to find a better pattern... build an easier algorithm from it... and do it as quickly as possible."- Genius))
What is this for? In the Russian Federation, imports are determined by entropy
like to learn to count by hand?
https://habrahabr.ru/post/171759/
Why do I need to know all this?
What is this for? In the Russian Federation, imports are determined by entropy
Like to learn to count by hand?
https://habrahabr.ru/post/171759/
Why the fuck do I know all this?
Do not worry.
Don't be shy.
Are you a boozer?
The algorithm for opening and closing an order is different. It is not clear what you find there through R, when you process only opening signals...
MSE can only give wrong measurements on check samples by reducing the value range from 100 to 85 on a constant basis.
Do not talk nonsense, 70% - an error from the wrong target, and these 17%, 30% - your fantasy numbers from the ceiling. There is a simple correlation between accuracy and sharp ratio, which after 55% gives cosmic values, you just do not understand what you are talking about 70%
I don't know what you're talking about either. Please explain:
1 What is this "not the right target"?
2. Reference to a "simple" relationship "between accuracy and sharp ratio"?
3. If your model provides Accuracy less than 0.75, send it to the furnace.
There are a lot of oracles on the branch, they say without proof of the experiment, without reference to third parties, just to make a fuss.
If you conducted your own experiments - give results, if there is a third-party study - give a link.
Good luck
For the first time in so many years I was banned. But I found out about it on the last day of the ban, so I planned to rest, so here I had to.
I'll jump right in on the topic. Wrong target. Even if it is not correct, the model will be optimized exactly to the target, even to incorrect...... After all, a wrong target is a wrong interpretation of it. Well, if the target is selected and the model is built on it, the model will meet the requirements of the target, and then what is it, right or wrong decides only the expert who has built it ... IMHO