Machine learning in trading: theory, models, practice and algo-trading - page 1914
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
I also smile when I read that someone has invented another grail)))) So good luck again)
There is a very interesting task - to forecast price movement using neural networks. This problem is solvable and there are already working solutions. Why does everybody focuses on getting profit? It is like learning to fly, but do not consider any flight until you reach the moon.
You cannot make a profitable trading neuronet out of nothing at once. First you have to learn how to predict something, then use this experience to understand the nature of the object you are investigating and learn how to work with it. Improve the network, raise the quality, and only then this quality will inevitably grow to successful trades.
Everyone is locked into deals, and there is no process of research and gradual growth. At every stage there has to be a result, a real prediction, even if it is of low quality. It is necessary to get a result, consolidate and then move on.
If you think in terms of profit/loss this problem will never be solved. You'd better give 100K for a sensible research with sensible explainable results.
Man, such interesting structures can be obtained in a compressed feature space, I don't know how to apply them, but it's fascinating.)
Two PCA features are presented in 3D umap space
==========
two indicators in three-dimensional space
))))
You know, maybe you should not look for such people on forums where stochastics optimize?
Maybe it is necessary to look for such people at specialized departments, and at several of them.
At one time people did not believe that it was possible to fly... The first flight happened and 86 years later we flew into space...
Here's how a robot trades with a knowledgeable person who has devoted his life to clever algorithms.
I could give you his contacts, but he doesn't need you, neither with his hundred, nor with a million, why? the answer is obvious and it is in the picture if you think about it
If you live on an island in a tribe and no one there can speak English, it's impossible to speak English, isn't it? )))))))
numbers on the screenshot - this is a measure of what?
The time of entry/exit, fin. res. may be something else, the white color of the font is the deal in the plus, orange in the minus.
entry/exit time, fin. res. may be something else, white color font deal in the plus, orange in the minus
see
neuronics?
Oh, don't you start with that...
ok mom's driver, i'm sick of you already sperdalai a lot
ahahahahah))))) there goes the shit out of all the cracks))))) the pictures they draw here ))))) clowns) my friend trades ahahahahahah))))))) keep predicting stochastics) dilettante
There is a very interesting task - to predict the price movement with the help of neural networks. This problem is solvable and there are already working solutions. Why does everybody cycles on making profit? It is like learning to fly, but do not consider any flight as a result until you reach the moon.
You cannot make a profitable trading neuronet out of nothing at once. First you have to learn how to predict something, then use this experience to understand the nature of the object you are investigating and learn how to work with it. Improve the network, raise the quality, and only then this quality will inevitably grow to successful trades.
Everyone is locked into deals, and there is no process of research and gradual growth. At each stage, there has to be a result, a real prediction, even if it is of low quality. It is necessary to get a result, consolidate and then move on.
If you think in terms of profit/loss this problem will never be solved. You would have been better to give 100K for a sensible study with intelligible explainable results.
I could put a million into the fund, no problem. Only it's like an experiment on psychics - what with 1966 1 million dollars to anyone who proves his psychic abilities.But so far no one has proved.There have been attempts, but alas. And the rest of the mass of psychics says that they can easily take this million, but they do not need it))) Quite like the one who posted above pictures of his friend - he does not need either hundreds or millions of dollars. This is an unpromising job at the moment.
I can put a million dollars into the fund, no problem. Only it's all like an experiment on psychics - that with 1966 1 million dollars to anyone who proves their psychic abilities.But so far no one has proved.There have been attempts, but alas. And the rest of the mass of psychics says that they can easily take this million, but they do not need it))) Quite like the one who posted above pictures of his friend - he does not need either hundreds or millions of dollars. This is an unpromising job at the moment.
got it
neuronics?
There are neuronics and adaptive filtering and spectral analysis and Q-learning and clustering and training with a teacher, in which the teacher himself, he pokes the points on the graph where you need to enter and punishes the robot himself when he makes a mistake. The robot is adaptive, it acquires new knowledge and forgets old ones if necessary, so it is a real beast.
I've had a neural network that predicts the movement of four pairs for four months now. It is of higher quality than any indicator, I demonstrate it publicly. I can not give the link, otherwise the admin will be hysterical again because there are a couple of paid services. But in any case it is a real neural network and it really sees and predicts something.