Machine learning in trading: theory, models, practice and algo-trading - page 2097
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
Ahahaha )) the Danish krone rules the euro )))
was just remembering... either burrito or anchovy... boruta, right
I was just remembering... it was either burrito or anchovy... boruta, right
I'm not impressed with it.
this hat should compress the feature space, so that the ns does not retrain to anything, to any noise
The rocket does the same thing, but without a neural network and all the convolution kernels are random. And then the best ones are chosen, by entropy or something else.
The problem with convolutional networks is the architecture selection, so use off-the-shelf models, all kind of restnets, etc.
Question for the newsmen, what happens on the 4th or 5th of each month?
Another question about scaffolding, can the target be set as "profit maximization" rather than class partitioning or regression?
With convolutional networks, the problem is in the selection of the architecture, so use ready-made models, all restnet, etc.
ready-made for what? you have to make your own, it's not that hard... it's harder to get started
Ready for what? You have to make your own, it's not that hard... it's harder to get started
Pre-trained. Mostly convolutional ones are used for picture recognition. Each layer highlights some features (stripes, corners), like in the brain. You can take a ready-made network (which is trained on supercomputers) and pre-train it on your examples.
Read more (from What Our Image Recognizer Learned)Pre-trained. Mostly convolutional ones are used for picture recognition. Each layer highlights some features (stripes, corners), like in the brain. You can take a ready-made network (which was trained on supercomputers) and pre-train it with your examples.
Read more (from What Our Image Recognizer Learnt)Did you understand what I just said? ) take a network trained on SEALs and retrain it on increments?
this eugenics I have not yet encountered
this hat should compress the feature space, so that ns does not retrain on anything, on any noise
The rocket does the same thing, but without a neural network and all the convolution kernels are random. And then the best ones are chosen, by entropy or whatever.
Try it, I'm not good at it.
Another question about forests, is it possible to set the target as "profit maximization" rather than splitting into classes or regression?
Profit maximization is an optimization task here other algorithms genetics, annealing...
Forces is teacher-assisted learning, you need partitioning...
I wonder how it all fits together.
did you understand what you suggested? ) Take a network trained on SEALs and train it in increments?
I've never seen such eugenics before
look at the link and browse below, you'll understand
Look at the link and scroll down below, you'll get the idea