Improving Machine Learning Ability with Fine-Tuning.
John Patrick Lalor, Hao Wu, Hong Yu · arXiv (Cornell University) · 2017
Item Response Theory (IRT) allows for measuring ability of Machine Learning models as compared to a human population. However, it is difficult to create a large dataset to train the ability of deep neural network models (DNNs). We propose fine-tuning as a new training process, where a model pre-trained on a large dataset is fine-tuned with a small supplemental training set. Our results show that fine-tuning can improve the ability of a state-of-the-art DNN model for Recognizing Textual Entailment tasks.