Python port of BYU CS 478 machine learning toolkit
Works with Python 2.7 or 3. Requires NumPy.
In order to use this toolkit, most commands will be similar to those given on the class website for the Java and C++ toolkits. With the assumption that you already have NumPy installed (see their website for installation instructions), usage is straight-forward.
As example, execute the following commands from the root directory of this repository.
mkdir datasets
wget http://axon.cs.byu.edu/~martinez/classes/478/stuff/iris.arff -P datasets/
python -m toolkit.manager -L baseline -A datasets/iris.arff -E training
Aside from needing to specify the module to run, commands follow the same syntax as the other toolkits.
For information on the expected syntax, run
python -m toolkit.manager --help
See the baseline_learner.py and its BaselineLearner
class for an example of
the format of the learner. In particular, new learners will need to override
the train()
and predict()
functions of the SupervisedLearner
base class.