Covers ARMA models for time series forecasting, discussing implications, properties of forecast error, challenges with predictions, and covariance models.
Covers the basics of text generation and the challenges of evaluating generated text using content overlap metrics, model-based metrics, and human evaluations.
