Machine learning in trading: theory, models, practice and algo-trading - page 2487
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 keep thinking about the web model (c59) (in the context of striving for balance/disbalance)... it's the mathematics of the model that scares me
By the way, yes, all models are usually levelled against reality by the fact that they all have some assumptions... as said above, it's better to go from the data to the model rather than from the model to build your analysis -- I sometimes miss this(the so-called "overlearning" state -- when you try to calculate an ordinary accounting balance from some ec. model)... Although, as always, the logic of buyer-seller interaction is closer to me (trivial, cheap/expensive - need/don't need - deficit/surplus - the most working model of trader and risk manager)... often quantitative assessments start to limp when differential equations are seen or due to imperfect statistical processing methods, to overcome which I have to wrap myself in decent additional statistical research/sample calculations (for the sake of historical output)
Can we already see the prototype of something?
Because all you talk about is just talking and quoting yourself, can we see at least a little bit of practice in this ocean of theory?
Can we already see a prototype of something?
Because all this is just talk and self-talk, can we see at least a little bit of practice in this ocean of theory?
Can we already see a prototype of something?
Because all you talk about is just talking and quoting yourself, can we see at least a little bit of practice in this ocean of theory?
What's your problem with translating theory into practice? - this is a rhetorical question, your answer does not even interest me (i already have to repeat it to you)... - why don't you understand from the 1st time? - another rhetorical question)... trolls are annoying!!! (going around shitting in a herd in the thread again)
Actually it is necessary to understand all philosophy of the field of IR, and without knowledge of a large volume of information in this field it is very difficult to come to philosophy of a question, and without it to come to correct conclusions!
I, for example, organized the process of obtaining a model for 2 hours, and a set of programs is selected and I do over and over the same actions. ONE AND THE SAME!!!! There is no need to bounce between optimization methods, network topology choices, or reshuffling of input data. The algorithm of model creation is rigid and cannot be changed, in other words, while you are still just looking, I have already found and stupidly use!!!!
and what is your problem with translating theory into practice? - this is a rhetorical question, your answer does not even interest me (I already have to repeat it to you)... - why don't you understand from the 1st time? - another rhetorical question)... trolls are annoying!!! (going around shitting in a herd in the thread again)
I'm not trolling, I want to understand you, do you have some kind of market outlook? a model in your head all this? do you realize that? what progress? maybe I want to learn from you, or to give you some advice ...
Or you just talk about everything from arbitrage and options to algotrading and IO, and you throw out links and self-citations...
While unfortunately I see only the second, and no one is interested and it makes no sense without practical experiment ...
So if I'm a troll in your paradigm, then you're a regular spammer)
Actually it is necessary to understand all philosophy of the field of IR, and without knowledge of a large volume of information in this field it is very difficult to come to philosophy of a question, and without it to come to correct conclusions!
I, for example, organized the process of obtaining a model for 2 hours, and a set of programs is selected and I do over and over the same actions. ONE AND THE SAME!!!! There is no need to bounce between optimization methods, network topology choices, or reshuffling of input data. The algorithm of model creation is rigid and cannot be changed, in other words while you are still just looking, I have already found and stupidly use!!!!
OK, show me the results
Actually it is necessary to understand all philosophy of the field of IR, and without knowledge of a large volume of information in this field it is very difficult to come to philosophy of a question, and without it to come to correct conclusions!
I, for example, organized the process of obtaining a model for 2 hours, and a set of programs is selected and I do over and over the same actions. ONE AND THE SAME!!!! There is no need to bounce between optimization methods, network topology choices, or reshuffling of input data. The algorithm of model creation is rigid and cannot be changed, in other words while you are still just looking, I have already found and stupidly use!!!!
Yep... but of course you can't test it ))))
And you've been using a ready-made rattle of Reshetov for years, from it you heard the word polynom, neuron, mgua etc. for the first time, but you claim to be a master of neural network training with 20 years of experience)) and at the same time a year ago asked to be taught to program in R or Python)))
Fools believe it, master, what the fuck to do).
A set of programs picked up and I do the same thing over and over again.
what? i don't know anything besides Python for MO (i haven't seen any special tools in c#, in c++ you have to write everything from scratch by all means), but even in Python please guide me on libraries, if possible... or what toolkit do you advise?
and how often and when do you use matrix calculus, if you do - are there libraries in nature so you don't have to write methods by hand?
p.s.
Aggregate profile of HFT bids presence fraction at best prices