Machine learning in trading: theory, models, practice and algo-trading - page 3241
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
imho, there will be no contest rather than vice versa..confirmation: in the most popular thread on the site about MoD - everything is still the same :-)
a question to im. drimer: where is the money, zin ?
There are no significant tangible results (except for personal experience and self-development) yet. And there is a significant risk that none will appear during the contest - and that would be a fiasco. The organisers have a difficult task to create the conditions for the appointed horses to at least reach the finish line.
ONNX could be an alternative for OpenCL. But this is just an idea for now.
No, it is not even close to being an alternative.
The main purpose of ONNX is as an open neuromodel exchange format and execution environment. Not training.
Will zipmap support be added? Not all models have it disabled when converting.
convenience for
ONNX: output parameter has unsuppotred type 'ONNX_TYPE_SEQUENCE'
Now if they go there, almost all of them will hit it, but they will not have the desire and ability to edit ONNX files.I can't say anything about zipmap yet. We need to fight the output
imho, there won't be a concours rather than vice versa...confirmation: in the most popular thread on the site about the MoD - everything is still the same :-)
question to im. drimer: where is the money, zin ?
There are no significant tangible results (except for personal experience and self-development) yet. And there is a significant risk that none will appear during the contest - and that would be a fiasco. The organisers have a difficult task to create the conditions for the appointed horses to at least reach the finish line.
No, it's not even close to being an alternative.
The main purpose of ONNX is as an open neuromodel exchange format and execution environment. Not training.
Yes, I understand, ONNX is a neural network storage format and graph computation.
If you get creative, you can get visual programming based on ONNX viewer.
PS not calling for anything, just thoughts.
There are no significant tangible results (except for personal experience and self-development) yet. And there is a significant risk that they will not appear during the competition and it will be a fiasco. The organisers have a difficult task - to create conditions for the nominated horses to at least reach the finish line.
That's roughly what I think. On a large interval tests can show profit, but also drawdowns up to half a year - a year. I.e. the model that happens to be not in the drawdown period in the month of the contest will win.
So you can easily put your perceived practical and practical result into the model and with a complete lack of competitors become a happy owner of 15 thousand USD.
And if you have a wife and mistress to take the entire prize )
in general it amazes the desire of offtoppers to assert themselves at the expense of the failures of others, apparently they do not have enough of their own because of an uninteresting existence :)I think about the same way. On a large interval, tests can show profit, but also drawdowns up to half a year - a year. I.e. the model that happens to be not in the drawdown period in the month of the contest will win.
And (un)by the way about the time factor...to show something about a strategy you need a quarter. Less than that will be a great random, even to watch will be uninteresting, more is hardly enough patience and attention of the public....
And now it's already September, and there is already December, where everything is completely different and the end of the year.
and (un)by the way about the time factor...to show something about strategy you need a quarter. Less than that is definitely a great random, even to watch will be uninteresting, more is unlikely to have the patience and attention of the public....
And now it's already September, and there is already December, where everything is completely different and the end of the year.
No, there's no point in watching less than a year...
But it makes sense to participate.The testing system will consist of three components:
Another question on MM, will there be any limitations here?