# Topics - Introduction - Affective computing in research and industry - Probability and linear algebra review - Extracting behavior 1: text - Machine learning 1: Linear and logistic regression - Machine learning 2: Gaussian Mixture Models - Extracting behavior 2: Audio - Machine learning 3: Hidden Markov Models 1 - Machine learning 4: Hidden Markov Models 2 - Hyperparameter optimization and regularization - Extracting behavior 3: Video - Deep learning 1: Feed-forward NN - Deep learning 2: CNN - Deep learning 3: LSTM/GRU - Deep learning 4: Transformers - Anomaly detection - How do we measure behavior? Designing datasets for affective computing. - Real world challenges - What can we do? Challenges and limitations - Sources of variation and methods to counter: audio - Sources of variation and methods to counter: text - Federated learning - Student-teacher models - Building models of a user - Detecting deviations from normal patterns - Visualizations - What should we do? Ethics, security, privacy (spotlight: HARPA) - Project presentations