Open sfvnielsen opened 7 years ago
Lets say we develop a DNN system that can detect propositions/statements better than the rulebased system. In that case it would be nice to analyse the examples that the rule-based systems misclassified and the DNN correctly classified
Machine Learning baseline: Logistic regression
We should have something to compare the performance of our network to.
What is the simplest we can think of that still makes sense?