I have an indicator written on PINESCRIPT and it is a machine learning logistic regression model.
I want it converted this indicator to python, then execute it as a live trading model in a way I can simply enter my API key and API secret and execute it on Binance via API.
The API library I would like to use for this is CCXT.
I also want the data streaming to be done via web sockets.
The pine script code is a logistic regression-based model with a gradient descent function for loss minimizing. In the actual code, the person has used a sigmoid function and not a logistic regression model. I need this sigmoid replaced with a proper logistic regression via scikit learn.
you must have the following skills for this
2) pine script
3) logistic regression - via scikit learn
4) Good understanding of gradient descent
5) Synthetic dataset to be run parallel to train the model while running
6) *important* I must be able to give multiple inputs in order to modify the model later. i.e logistic regression function to be generated via a data frame with multiple columns of inputs and the number of columns must be able to increase or decrease based on how many inputs i want to test the model on.
8) parallel processing
9) proven experience in coding a model that can be executed live
10) access to your own tradingview terminal so I can give you the name of the indicator that is available to all public
11) The tradingview model has a repainting issue. I would like the python code to generate the entry and exit signal once the BAR is actually closed and on the open of the next bar. Essentially it creates a lag.
12) I would like an additional part added to the exit signals as a trailing stop based on a dollar amount I specify AFTER the exit signal is triggered from this model.
Please bid only if you can clearly guarantee me you can fulfill all criteria above.