Ren, Jason J.2020-08-282020-052020-06-172020Ren, Jason J. 2020. Prediction Focused Topic Models via Feature Selection. Bachelor's thesis, Harvard College.https://nrs.harvard.edu/URN-3:HUL.INSTREPOS:37364737Supervised topic models are often sought to balance prediction quality and interpretability. However, when models are inevitably misspecified, standard approaches rarely deliver on both. We introduce a novel approach, the prediction-focused topic model, that uses the supervisory signal to retain only vocabulary terms that improve, or at least do not hinder, prediction performance. By removing terms with irrelevant signal, the topic model is able to learn task-relevant, coherent topics. We demonstrate on several data sets that compared to existing approaches, prediction-focused topic models learn much more coherent topics while maintaining competitive predictions.application/pdfenThesis or Dissertation2020-08-28