The project is to help me understand and create a randomForest machine learning model in R.
I have data in an xts that looks like this: http://pastie.org/7819102
(please click expand the view to view it more easily)
the data is 90 000 rows.
Plotted it looks like this: http://i.imgur.com/KUskvIh.png
This is a financial timeseries.
I want to create a randomForest machine-learning model that takes into account the other features in the data, like smafreq & atr to determine when it is a good time to sell or buy if the Close column is outside the 2 standard deviation bands in the graph.
This might require some feature extraction and engineering to order the features in formats that the R randomForest function can analyse. Please propose what your first step would be.
The idea is to create a model and at the same time help me understand it so I can create new models based on new features in the future.