AI stock market recommendation generator.
- The tool target is to provide insight of stock market (NASDAQ for now) in form of a list that describes possible stock value change after specific time.
- User should provide time intervals that should be analyzed and time units in future to get prediction for. (Short-Term Swing Trading for now)
- Tool gives a list of analysed stocks, predicted price movement and probability of analysis being right or wrong.
DISCLAIMER: Make trades at your own risk. Decisions should be your own!
- Install requirements.txt
Use: python tradinga.py
and add arguments.
- Train model by using
train
argument. - Test model reliability by using
metrics
argument. Use different symbols to test. Tweaksettings.py
for best results. - Either use
predict
argument to predict specific symbol value afer provided date or usepredict_market
argument to get summarized report of all market of downloaded symbols.
Note that predict_market
is time consuming action!
train
Create or retrain model (While training process can be stopped with Ctrl+C)--count COUNT
Count on how many symbols model has to be trained (Default: All symbols)-c
Clean settings and delete model before training
metrics
Get metrics for symbol and strategy-s/--symbol SIMBOL
Single symbol analysis (REQUIRED)-p
Plot data of predicted values on top of actual values
predict
Predict symbol value-s/--symbol SIMBOL
Single symbol analysis (REQUIRED)--last-date YYYY/MM/DD
Make prediction from this date
predict_market
Predict next market values and make csv with predictions--last-date YYYY/MM/DD
Make prediction from this date--count COUNT
How much symbols to analyze. Will go in random order (Default: All symbols)
Code consists of different classes. This graph should approximately represent data flow within:
GRAPH TO BE ADDED
To make sure issues are closed automatically when iteration is closed, changes goes to main, follow this guide: