Download
Deep learning models have achieved impressive prediction performance but often sacrifice interpretability, a critical consideration in high-stakes domains such as healthcare or policymaking. In contrast, generalized additive models (GAMs) can maintain interpretability but often suffer from poor prediction performance due to their inability to effectively capture feature interactions. In this work, we aim to bridge this gap by using pre-trained neural language models to extract embeddings for each input before learning a linear model in the embedding space. The final model (which we call Emb-GAM) is a transparent, linear function of its input features and feature interactions. Leveraging the language model allows Emb-GAM to learn far fewer linear coefficients, model larger interactions, and generalize well to novel inputs. Across a variety of natural-language-processing datasets, Emb-GAM achieves strong prediction performance without sacrificing interpretability.
Features
- For finer control, you can instead clone and install this repo from source
- Interpretable linear model that leverages a pre-trained language model to better learn interactions
- One-line fit function
- Emb-GAM uses a pre-trained language model to extract features from text data
- Combines features in order to extract out a simple, linear model
- The best way to use Emb-GAM is through the imodelsx package
- Emb-GAM achieves strong prediction performance without sacrificing interpretability
Project Samples
