BoostedBuidlingProcessForADG

Implementation of A novel machine learning method for my researches in this field.

This method is only for binary classification tasks up until current version.

Quick Start

install

pip install bbp4adg

import and initialize:

# from exports import BBP, ADG
from bbp4adg import BBP, ADG

adg = BBP(threshold=delta)

start training

# Or adg = ADG(threshold=delta) for original adg
model, perf =adg.fit(X,y)
print('arguments count:', len(model.arguments)) 
print(model.arguments)
print('relations count:', len(model.relations))
print(model.relations)
print("accuracy:", perf)

evaluate on test set

print("test accuracy:", adg.score(X_test,y_test))

ToDos

1. more comments
2. doc tests
3. more semantics than grounded

How to contribute

Issues and PRs are welcomed.

Please read the contributing document.