Machine learning in trading: theory, models, practice and algo-trading - page 2617

New comment
 
mytarmailS #:
I don't write often either, I read more often...
I simply search for a site and read questions and answers from people, and not only this site, there are also purely on quanta


I went to quant.stackexchange for example

you type neural network into the search engine

andread questions about neural networks and smart answers.


I do not get stupid rubbish, everything is clear and to the point, for unintelligent writing they give you a "minus", it is the ideal!

It would be the same here if it wasn't for the Market. Developers have gone there, it's not profitable for them to share and write, but they like to read ) The forum is now more of an appendix to the technical part of the platform, and as for TC, it's in the Marketplace.
 
Maxim Dmitrievsky #:
It would be the same here if it wasn't for the Market. Developers went there, it's not profitable for them to share and write, but they like to read ) Now the forum is more of an appendix to the technical part of the platform, the trading part is in the Marketplace.
Yes... They love to read. They like to take but not to share
 
mytarmailS #:
Yes... They love to read. They like to take but not to share.
For example, you write an article - a week later, 5-10 products appeared in the market, almost entirely copied. You got $200 for it, they earned a few thousand dollars each. The motivation to share is slightly diminished, but the enthusiasm is there 😀
 
So the logical move would be to reverse-engineer the TS from Market, including with the help of MO. Because the algorithm base there is already huge. That's roughly where the future lies. I'm guessing. It's like in music and creativity, there's oversaturation, then a bunch of remakes and remixes.
 
Maxim Dmitrievsky #:
So the logical move would be to reverse-engineer the TS from Market, including with the help of the MO.

It's not that simple...

Profitable strategies don't trade in a moving window, so you can't simulate them with MO because by standard AMOs work with tabular data, and tabular data is essentially a calculation of stuff in a moving window...


Here's an example from the ceiling: Let's call this "Profitable Strategy": Wait for breakout of weekly low, then go back and wait for some kind of candlestick configuration - enter...

How can you find such a pattern in the MO if you have table data, i.e. you look for the last n candlesticks, the answer is nothing.

Of course you can create traits for this "Profitable Strategy" to make it work, but you have to know this strategy to create traits for it and we don't know it...


There are only two algorithms that can solve these problems, maybe only one... But there is

 
a cool python backtester has turned up I want one for my rca))
Backtesting.py - Backtest trading strategies in Python
  • reviews: 1
  • kernc.github.io
Fast Python framework for backtesting trading and investment strategies on historical candlestick data.
 
mytarmailS #:

It's not that simple...

Profitable strategies don't trade in a moving window, so you can't simulate them with MO because by standard AMOs work with tabular data, and tabular data is essentially a calculation of stuff in a moving window...


Here's an example from the ceiling: Suppose it's a "Profitable Strategy": Wait for a breakout of the weekly low, then go back and wait for a candle configuration - enter...

How can you find such a pattern in the MO if you have table data, i.e. you look for the last n candlesticks, the answer is nothing.

Of course you can create traits for this "Profitable Strategy" to make it work, but you have to know this strategy to create traits for it and we don't know it...


There are only two algorithms that can solve these problems, maybe only one... But there is

It depends on the strategy. If external sources like news are used, then yes, it is more difficult. If solely based on price, then it's probably a matter of time
 
mytarmailS #:
a cool python backtester has turned up - i want one for my rca))
Been there, it's slow.
 
Maxim Dmitrievsky #:
It depends on the strategy. If you use external sources like news, then yes, it's harder. If solely based on price, then it's probably a matter of time.

The way everyone uses the MO, it's just a dumb memory tool that looks at the last 5-10 candles, and it can't take anything from this data, that's a fact

 
mytarmailS #:

The way everyone uses the MO, it's just a dumb memory tool that looks at the last 5-10 candles, and it can't take anything from this data, that's a fact

Let's check it out, I've been wanting to do that for a long time.)
1...261026112612261326142615261626172618261926202621262226232624...3577
New comment