# Machine Learning

## Logistics

Location | ESJ 2212 | |

Time | Mon/Wed 15:30 - 16:45 | |

Webpage | http://umiacs.umd.edu/~jbg/teaching/CMSC_726/ | |

Mailing List | Piazza | |

Required Text | Understanding Machine Learning: From Theory to Algorithms (UML) | |

Suggested Text | Foundations of Machine Learning (FML) | |

Suggested Text | Machine Learning: A Probabilistic Perspective (Murphy) | |

Syllabus | https://docs.google.com/document/d/1JyKb5JolsvttYb5jem1CgNskDUT79_yYVd5zKTTTg9s | |

Video Lectures | YouTube | |

Grades and Submission | ELMS |

## People

### Professor

Jordan Boyd-Graber

AVW 3155

Office Hours: Starting 30. August, Wed 14:00 - 15:00 and by appointment

### TA

- Fenfei Guo
(Despite Colorado URL, will be PhD student at UMD in Fall) Office
hours Thu 14:00 - 16:00 in AVW 3164
## Schedule

Date In-Class Topic Assignment Due Lecture Mon 28. Aug 1. Machine Learning as a Black Box [Video A B C] [PDF A B C] **Readings:**- UML 1
- (Assumed background) UML Appendices B-C
- (Assumed background) Probability (Chapters 1-2)
- (Assumed background) Joint Probabilities and MLE

Wed 30. Aug 2. Logistic Regression [Video A B C] [PDF A B C] **Readings:****Optional:**- Notes on Logistic Regression (discusses updates)
- Ng and Jordan proofs about large data equivalence (only read this if you thought that the Mitchell reading wasn't theoretical enough)

Mon 4. Sep Labor Day Wed 6. Sep 3. Stochastic Gradient Optimization for Logistic Regression [Video A B] [PDF A B C] **Readings:**- Bob Carpenter on fast regularized updates (helpful for homework)

Fri 8. Sep Homework 1 Due K Nearest Neighbors Mon 11. Sep LAB DAY for HW2 Wed 13. Sep 4. Feature Engineering [Video A B] [POS Script] [data] [Slides] Fri 15. Sep Homework 2 Due Logistic Regression **Readings:**Mon 18. Sep 5. PAC Learnability [Video A B] [PDF A B] **Readings: (Choose one)**- Computational Learning Theory
- UML 4-5
- FML 2

Wed 20. Sep 6. VC / Rademacher Complexity [Video A B] [PDF A] [PDF B] [PDF C] **Readings: (Choose one)**- FML 3
- UML 6, 26

Mon 25. Sep Class Cancelled Wed 27. Sep 8. Support Vector Machines [Video A B C D] [PDF A B C D] **Readings: (Choose One)**- UML 15-16
- FML 4-5

Fri 29. Sep Homework 3 Due Feature Engineering Mon 2. Oct 9. Boosting [Video A B] [PDF A B] **Readings:**(Choose one)- UML 10
- FML 6

Wed 4. Oct 10. Regression [Video A B] [PDF AB] [Data] **Readings:**(Choose one)- FML 10
- UML 9.2

Fri 6. Oct Homework 4 Due Learnability Mon 9. Oct 11. Structured Perceptron [Video A B C D E F] [PDF A B C D] **Readings:**(Choose one)Wed 11. Oct 12. Loss Functions and Multilayer Backprop [PDF A B C] [Video A B1 B2 C] **Readings:**(Choose one)- UML 20

Fri 13. Oct Homework 5 Due SVM Mon 16. Oct Review / Catchup Wed 18. Oct Midterm Fri 20. Oct Project Milestone Project Proposal Mon 23. Oct 13. Representation Learning [Video A B C] [PDF A B C D] [Exam Questions Questions Ex] **Readings:**Wed 25. Oct 14. K-Means / Mixture Models [Video A] [PDF A B] [Questions Ex] **Readings:**- CIML chapter 15.

Mon 30. Oct 15. Dirichlet Process / Gibbs Sampling [Video A] [PDF A B] [In Class ] **Readings:****Optional:**Wed 1. Nov 16. Topic Models [Video A] [PDF A B] [Questions Ex **Readings:**- Chapter 1 of Applications of Topic Models

**Optional:**Mon 6. Nov 17. Variational Inference [Video A] [PDF A B] [Questions ] **Readings:**Wed 8. Nov 18. Variational Autoencoders and Generative Adversarial Networks [PDF A B C D] [Video Admin A B] **Readings:**Fri 10. Nov *Project Milestone*First Deliverable Mon 13. Nov 19. Memory Models (LSTMs, GRUs) [Video A B] [PDF A B C D] [Ex] **Readings:**Wed 15. Nov 20. Reinforcement Learning [Video A B C D] [PDF A B C D] **Readings:**Fri 17. Nov Homework 6 Due Variational Mon 20. Nov 21. Ranking, Regret, and Multiclass [PDF A B C D] [Video A B C] **Reading:**Wed 22. Nov Thanksgiving Mon 27. Nov 22. Fairness, Acountability, and Transparency [PDF A B C] [Video A B C] **Reading:**Wed 29. Nov 23. Will AI kill and/or enslave humanity? [PDF] [Video] **Reading:**Mon 4. Dec 24. The Culture of Machine Learning [PDF] [Video] **Optional:**Wed 6. Dec Class Cancelled (NIPS): Come to class time to review for final Fri 8. Dec Homework 7 Due Deep Learning Mon 11. Dec Midterm Fri 15. Dec, **1:30**Final Presentations Final Report