Pythonic Tutorial 17-18

Pythonic on GitHub
Pythonic Tutorial Part 1-6
Pythonic Tutorial Part 7-11
Pythonic Tutorial Part 12-16

17. Support Vector Machine

The Support Vector Machine element let’s you do a binary classification of the input data. The input must be a tuple of sample data and (binary) features ([0,1,1, 0, …]). Based on the selected ratio, the data is broken up into training and evaluation data. The output of this element is a contingency table.

To improve the prediction, the input data should be centered and/or scaled.

You can define a file path (absolute or relative to $HOME) where the resulting model is saved (it’s a pickled Scikit Learn SVC).

Scikit Learn Support Vector Machine

18. Support Vector Machine Prediction

This element creates a prediction based on the input data and a model. You can specify the absolute or relative (to $HOME) file path to the model. If the input data is a list, you can activate that only the last value within this list gets predicted. This will reduce the CPU utilization if you want to pass in OHLC data regularly.

The output of this element is a Python list of
binary classification ([0, 0, 1, 0, 1, 1] — with just one element when Predict only last value is activated.)