Plan is to build a functional for trading btc/ethereum using the gdax-api.
Historical btc data is available in batch from http://api.bitcoincharts.com/v1/csv/
#1 Create a API to consume GDAX streaming API and save the data to a postgres database.
#2 Develop a ML model using technical analysis to perform next period price prediction.
*Use Keras to prototype a Convolutional Bidirectional Recurrnet NN on different aggregations of data to calcualte OHLC. (5/10min(?), 1hr, 1d(30), 1w(26), 1m(24))
Current raw data should exist as: (symbol), price, volume, timestamp
*Create standard technical analysis features. SMA/EMA/BollingerBands/MACD at different time points for compairison purposes
#3 Deploy ML model to score on streaming API every X minutes and perform automated trading using the stream taking into account current position.
REFERENCES:
http://pandas.pydata.org/pandas-docs/stable/timeseries.html#offset-aliases https://stackoverflow.com/questions/36222928/pandas-ohlc-aggregation-on-ohlc-data
Misc:
https://stackoverflow.com/questions/46070126/google-finance-json-stock-quote-stopped-working/ http://www.marketcalls.in/database/google-realtime-intraday-backfill-data.html