Machine learning in trading: theory, models, practice and algo-trading - page 1883
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
There's not going to be a movie, let's split up. Box and Jenkins rejoice
The results of this study suggest that simple classical methods, such as linear methods and exponential smoothing, outperform complex and sophisticated methods, such as decision trees, Multilayer Perceptrons (MLP), and Long Short-Term Memory (LSTM) network models.
There's not going to be a movie, let's split up. Box and Jenkins rejoice
The results of this study suggest that simple classical methods, such as linear methods and exponential smoothing, outperform complex and sophisticated methods, such as decision trees, Multilayer Perceptrons (MLP), and Long Short-Term Memory (LSTM) network models.
an article by imbeciles for imbeciles
By the way, maybe the experts can help. Here's a question:
The task, for example, to distinguish a cat from a dog by photo. What is the right way to learn?
1. Show pictures of cats and dogs only, i.e. binary classification.
2. Separately teach only cats and "not cats" (protos chaos) + separately dogs and "not dogs" too, i.e. two cycles of training and two models at the output.
3. Make a classification of three - cats, dogs and chaos. I.e. there will be one model, but the answer is a classification of three options.
Right now I have the first option, and it's clearly crooked. The problem is that neuro learns only one of the options well, conditionally sees only "cats" well, and recognizes dogs poorly. Example, on backtests the models are good at detecting upward price movement and ignore downward movement. If the upward guess reaches 67%, the same model guesses downward only 55%. The "up" and "down" from model to model can change places.
Usually they say try everything and pick the best one. Imho, 2 grids, each working with its own task (cats or dogs), plus set a threshold on the output. Then you need to somehow determine which examples are recognized the worst, they need to be processed or additional training. At the input serve those examples that will be used in practice, there is no point in showing the house in training, if it will not be in the test. Also, the pictures on the trainee and the test should be processed equally.
By the way, I am open for partnership if there is any resource, not necessarily a material one - an audience for promotion or an opportunity to organize further research based on what you already have. Just on a crank, this topic to a good level can not raise. You need real experts from different fields.
Throw in your profile link to telegram-channel. It would be interesting to see.
Usually they say try everything and choose the best option. Imho, 2 grids, each working with a different task (cats or dogs), plus at the output set the threshold. Then you need to somehow determine which examples are recognized the worst, they need to be processed or additional training. At the input serve those examples that will be used in practice, there is no point in showing the house in training, if it will not be in the test. Also, pictures on the tray and on the test should be processed equally.
Why view the price through the prism of a picture if you have the exact coordinates. The price and the time. It remains to describe the model you are interested in. And since there are a lot of them, then selectively. The machine recognizes them 100% even without training.
Throw a link to your telegram channel in your profile. It is interesting to watch.
signals from indicators on pause till monday, go to "History" button to see how they look like and neuro signals
Usually they say 1. Try everything and choose the best option. Imho, 2 . each grid works with a different task (cats or dogs), plus on the output set the threshold. Then you need to somehow determine which examples are recognized the worst, they need to be processed or additional training. At input 3. serve those examples that will be used in practice, there is no sense to show the house in the training, if it will not be in the test. Also pictures on the traine and the test should be equally processed.
2. Tried it, it got worse.
3. How do you know which ones she's using? It's a black box, so I feed everything.
Why look at the price through the prism of a picture if there are exact coordinates. Price and time. It remains to describe the model of interest. And since there are a lot of them, selectively. The machine recognizes them 100% even without training.
The question was about cats and dogs. In the same way it is possible to teach the network to perform a graphic analysis.
Why look at the price through the prism of a picture if there are exact coordinates. Price and time. It remains to describe the model of interest. And since there are a lot of them, selectively. The machine recognizes them 100% even without training.
The question was about cats and dogs. And so you can teach the network graphical analysis to try
Maybe you can show the network screenshots, but after optimization you will come to Open, Close, High, Low