Machine learning in trading: theory, models, practice and algo-trading - page 912
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 do not have an equal number, because the TC does not assume profit on every sneeze. In the file 1-2 column information, 3 and 4 are two independent targets, and the other predictors for them.
This is bullshit. Are you really feeding full minutes into the NS. I feel bad for you. You don't understand the nature of neural networks. It's not a dumpster to throw HUNDREDS of garbage in. I know. Upload the data for the last month and no more, since you can't part with the minutiae, and then we'll see what's up ...
This is bullshit. Are you really chowing down completely on minutes in the NS. I feel bad for you. You don't understand neural networks. It's not a dumpster to throw HUNDREDS of garbage in. I know. Upload the data for the last month and no more, since you can't part with the minutiae, and then we'll see what's up...
I see, your system is for fitting on small time intervals - I'm looking for stable systems on a long period of time....
Thank you.
I change the training parameters of the standard BP according to the results of the previous training. In essence - it is the same annealing, but controlled manually.
I don't use NS R packages, but I don't rule them out in the future. I work with a different environment.
There are genetic, evolutionary, Bayesian methods to obtain optimal hyperparameters of models. There is no chance to do it by hand.
IMHO, of course.
Good luck,
A couple of days ago I talked to Doc and told him how I was evaluating the obtained model. I told him right away that the model is retrained and put forward the theory that if the model is retrained, then we should invert the signals and earn money. He was skeptical, and I myself as well do not take this approach well. But the reality tells about it. Before the blue line the period of training, after today and what do we see????
So why, knowing that the model is retrained before the opening of trading today, I did not turn it? I'm asking you ..... I have come to the market to earn, not to assert myself, and it does not matter that the earnings go against my logic or established thoughts......
Therefore, approximately by September I plan to record a video, which will be interesting not only for beginners, but also for the experienced, where I will tell a number of reasons why you do not succeed.........
I see, your system is for small timeframe fits - I'm looking for stable systems over a long period of time....
Sorry, there are no such systems, otherwise people would have been using it long ago and the market would have lost its effectiveness. Long term systems can usually afford to go into drawdown and there is no guarantee that it will come out of it. I'd like to see you at that time, wondering if it will or won't. Bullshit. The market is in the moment. Here and now.... and mistakes on it are unacceptable....!!!!
A couple of days ago I talked to Doc and told him how I was evaluating the obtained model. I told him right away that the model is retrained and put forward the theory that if the model is retrained, then we should invert the signals and earn money. He was skeptical, and I myself as well do not take this approach well. But the reality tells about it. Before the blue line the period of training, after today and what do we see????
So why, knowing that the model is retrained even before the opening of today's trading, I did not turn it? I'm asking you ..... I came to the market to earn, not to assert myself, and it does not matter that the earnings go against my logic or established thoughts......
Therefore, around September I plan to record a video, which will be interesting not only for beginners, but also for veterans, where I will tell a number of reasons why you can not do it.........
learn how to do it properly
Sorry, but there's no such thing, otherwise people would have used it a long time ago and the market would have lost its effectiveness. Long-playing systems can usually afford to go into drawdown and there is no guarantee that it will come out of it. I'd like to see you at that time, wondering if it will or won't. Bullshit. The market is in the moment. Here and now.... and mistakes are unacceptable....!!!!
It follows from these words that training on a large amount of data will give bad results because the market is changing, but then how does one manage to train a primitive tree to create rules for this huge sample with a small margin of error? Randomness?
Sorry, but there's no such thing, otherwise people would have used it a long time ago and the market would have lost its effectiveness. Long-playing systems can usually afford to go into drawdown and there is no guarantee that it will come out of it. I'd like to see you at that time, wondering if it's going to come out or not. Bullshit. The market is in the moment. Here and now.... and mistakes are unacceptable....!!!!
The market is a daily job and if I have to build models every day (which I did at one time) I will continue to do so, because I came here to make money, not to show off. Beginners and slow learners at first try to make their views reality, and you should on the contrary change their views in accordance with reality. This is one of the reasons why you do not succeed. It's just that I can explain it much more clearly in a video, rather than write all this stuff... Maybe the printed word doesn't make sense to you. Well... I'll speak in my voice :-)
Learn how to do it right.
Cool.
There are genetic, evolutionary, and Bayesian methods for obtaining optimal hyperparameters of models. It is not promising to do this hand-to-hand.
IMHO, of course.
Good luck
Annealing is annealing in Africa.) For effective annealing, the main thing is to move to the next step (step reduction) after the previous one is done. Otherwise, we may get somewhere in the local min-max, and further it will not come out.
I absolutely disagree with you on this point.