Accounting for hidden common causes when inferring cause and effect from observational data. Learning Patient Representations from Text. Learning Neural Emotion Analysis from 100 Observations: The Surprising Effectiveness of Pre-Trained Word Representations. Determining the best classifier for predicting the value of a boolean field on a blood dono …