Machine learning in trading: theory, models, practice and algo-trading - page 3127
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
It's the Garchies again.
What's wrong with garches?
Only two options, trend trading doesn't count, so it's either patterns via MO or garchi.
What are we trading?
Trends?
Patterns?
Statistics on increments through garci?
You have to decide first, then draw conclusions.
Do you have a question about sample size?
I don't know what to answer in essence.
A pattern of the beginning of an intraday trend, probably.I wonder what kind of library this is?
To answer your question: This DLL is mine for checking access keys from the real and demo accounts of my MarketTrader Expert Advisor users. The connection goes to my cloud in Google Cloud. The DLL is also used for automatic periodic optimisation. In the attachments to posts DEMO source (part of the code is not for publication - in particular the code of the auto-optimiser).
Hi!
Sometimes I look here optimised charts for 10-20 years. Nor in some places where in 3 years where in 2 years where in 1 one year - there are plums. But the chart is growing.
Is it possible to find patterns of growth spikes, in which the robot trades profitably?
I've noticed an interesting thing here.
Everyone knows about data drift. We are used to kicking only predictors, but I decided to see what happens to the strategy itself over time.
I don't understand what this experiment is about. The elasticity of strategy outcomes depends on the drift of the predictors, no? The strategy depends on the predictors.
I don't understand what this experiment is about. The elasticity of strategy outcomes depends on the drift of predictors, no? Strategy depends on the predictors.
It's about the drift of the target was in fact - changes in its statistical properties over a time interval...
It's about the drift of the target was essentially - changes in its statistical properties over a time interval...
which depends on feature drift
if the dependence curve is plotted against the hypothetical-theoretical drift of traits (CATE among housewives), you will get the elasticity of change in the properties of the target.what depends on the drift of traits
If you plot the dependence curve against the hypothesis-theoretic drift of traits (CATE among housewives), you get the elasticity of change of the target propertiesUnfortunately, I don't understand your answer. What exactly do you need to do? And what does it mean " hyptoetic-theoretic drift of traits" - is it from all or from each one individually? Have you tried doing this in python? For A/B testing we know the transition point, but here there is no such point - gradual change.
The drift of predictors can be manifested by a change in the variance and a change in the probability distribution.
Accordingly, in the first case, the logic remains in the past sample, but the greed method simply cannot pull it out.
In the second case, the logic of the consequences of the events described by the predictors has changed.
Separately, we can note the drift caused by the lack of data normalisation. In our case, it is relevant when the price moves out of the ranges and the predictors do not take it into account. For example, measuring something only in points.
Means the deck's been shuffled badly
bias - variance tradeoff