WarrenBuffet - predicts stock movement
by 
DataJhoom

src - contains scala code for TwitterStream, Model and Predict components
script - contains bash cmds to invoke the 3 functionality
ui - contains web related code for displaying stock recommendations

For every company, a model is created
data - contains sample format to feed ML Lib 
APPL.data accumulates <label> List<features> over a sliding window time period and this data is fed to generate/update the model
APPL_last.data accummulates List<features> over a shorter window and uses the model to predict the label which is sent to the UI

Here label is the classifier used to predict hold/sell/buy