Schedule is tentative. Check regularly for updates. Readings beginning with “E” are from Eisenstein’s Natural Language Processing, while readings beginning with “JM” are from Jurafsky and Martin’s Speech and Language Processing.
| DATE | TOPIC | READINGS | ASSIGNMENT | LINKS |
| 2/2 | Two Cultures of Linguistics | E1 | ||
| 2/4 | Naive Bayes and Generative and Discriminative Models | E2.1-2.2;4.3-4.4; JM4 | HW1 out; PA1 out 2/5 | |
| 2/9 | Logistic Regression | E2.5-2.7; JM5 | ||
| 2/11 | Stochastic Gradient Descent | |||
| 2/16 | Perceptrons | E2.3-2.4;2.8 | HW1 Due | |
| 2/18 | Multilayer Perceptrons | E3.1-3.3; JM7 | PA1 Due 2/19; PA2 out 2/19 | |
| 2/23 | Computation Graphs | |||
| 2/25 | More Computation Graphs | |||
| 3/2 | N-grams and Vector Semantics | E6.1-6.2;6.4-6.5; JM3 | ||
| 3/4 | Distributional Semantics and Word Vectors | E14; JM6 | PA2 Due 3/5 | |
| 3/11 | Singular Value Decomposition | PA3 out 3/12 | ||
| 3/16 | Hidden Markov Models | E7.1-7.4;8.1; JM8;A | ||
| 3/18 | Conditional Random Fields | E7.5 | ||
| 3/23 | Structured Perceptrons and Review of Neural Networks | Bengio 1999; Collins 2002 | ||
| 3/25 | Recurrent Neural Networks | E6.3;7.6; JM9 | HW2 out; PA3 Due 3/26 | |
| 3/30 | Fancy RNNs | |||
| 4/1 | Syntactic Theory (Constituent) | E9.2; JM12 | PA4 out | |
| 4/6 | Syntax (Part 2) | HW2 Due | ||
| 4/8 | Earley Algorithm | |||
| 4/13 | CKY Algorithm | E10; JM13;C | HW3 out | |
| 4/15 | CKY Algorithm (Part 2) | PA4 Due 4/16; PA5 out 4/16 | ||
| 4/22 | Dependency Parsing | E11; JM14 | HW3 Due 4/23 | displaCy |
| 4/27 | Machine Translation | E18; JM11 | ||
| 4/29 | Attention | |||
| 5/4 | Transformers | PA5 Due; HW4 out; HW4 Due 5/14 |