Automated Reasoning on Machine Learning Model of Legislative Election Prediction
Yanti Rusmawati · 2021
Prediction models using machine learning have been utilized in various fields, including the general election prediction model. However, we still need more insight into the model result through explainable AI. To reason the result, in this in-use paper, here we compare two approaches: using ontology reasoner Protégé and Silas (a machine learning tool empowered with automated reasoning). Using the data set of the Indonesia legislative election in 2019, we build the prediction model, followed by extracting the formula from the decision tree then reasoning the model predicates. The result shows that to some extent we can have a better understanding of the reasonable result from the machine learning prediction model.