Open-source LSTM chain for quant

What the code is for ? This is an open source chain for testing trading opportunities, it allows to:

• Get financial information from Quandl, and define wanted financial technical indicators,
• Set up a LSTM model to "learn the knowledge" from the training set,
• Then make prediction in order to provide trading signals for the test set,
• And run Backtrader to test the strategy performance. The chain includes the following modulus:

1. Main.py is the main code to run the different modules;
2. A_Parameters.py defines the parameters (buy/short threshold, superparameters, train/test set etc.) for iterate testings;
3. B_SourceData.py defines the sources of financial information and technical indicators, needed for the machine learning model;
4. C_StructureData.py is the module allowing restructuring the data, defining the variables and labels for the supervised learning;
5. D_ModelTrain.py defines the LSTM model architecture, and runs the training process;
6. E_PredicAll.py allows to make prediction for the test data set;
7. F_BackTesting.py allows to run the backtrader program, and provide backtesting results for the training and testing data sets.

How to run the code ? In order the run this chain, you need to:

• get an Quandl key and put it into the B_SourceData.py,
• set up python 3.5 environment
• install libraries (keras, numpy, pandas, quandl, talib, matplotlib, sklearn, datetime, backtrader)

How to contribute ? To indicate....