Deep Learning - CPGEI

General Information

Objective: Study of the main deep learning methods and their applications.

Syllabus: Machine learning basics; Deep Feedforward Networks; Convolutional Neural Networks (CNN), Recurrent Neural Networks (including Long-Short-Term Memory Networks - LSTM); Generative Adversarial Networks (GAN); Object Recognition (including You Only Look Once - YOLO); Natural Language Processing (NLP); Advanced models; Frameworks for deep learning and Practical Aspects; Applications of deep learning models for real-world problems.

Duration/credits: 45 hours/3 credits (12 weeks).

Time: Tuesdays (17:50 - 21:10) - online with synchronous (SYN) and asynchronous (ASYN) classes.

Grade: Assignments (40%) and Final Project (60%).

Lecturers: Heitor Silvério Lopes and André Eugenio Lazzaretti.

Collaborators (PhD students): Andrei Inácio, Matheus Gutoski, Anderson Brilhador, and Clayton Kossoski.

Bibliography and Support Materials

Books:

Other Courses:

Online book:

Misc:

Week 1 - September 28

Lecturer:

  • Heitor e André.

Content:

  • (SYNC) General overview and Rules (slides, video).

  • (SYNC) Introduction to Deep Learning (slides, video).

  • (ASYNC) Python Basics (Lecture 3 and 4 from this link; Lectures 1.2, 1.3, 1.4, 1.5, 2.1, 2.2, 2.3, 2.4, and 2.5 from this link).

  • (ASYNC) Linear Algebra Review (slides, video).

  • (ASYNC) Statistics Review (slides, video).

  • (ASYNC) Numerical Computation (slides, video).

Assignments:

  • No assignments.

Week 2 - October 05

Lecturer:

  • André e Matheus.

Content:

Assignments:

  • (SYNC) Machine Learning Basics (link1 and link2) (Python notebook - download and edit your own code).

    • Deadline: October 08.

Week 3 - October 12 - Holiday or Recess

Week 4 - October 19

Lecturer:

  • André e Matheus.

Content:

Assignments:

  • (SYNC) Deep Feedforward Networks (link).

    • MLP example using Keras (link).

    • Tutorial - Titanic Challenge (link).

    • Leaderboard (link).

    • Deadline: October 28

Week 5 - October 26

Lecturer:

  • André e Matheus.

Content:

  • (ASYNC) Convolutional Networks (slides, video).

  • (ASYNC) CNN Architectures (video).

Assignments:

  • (SYNC) Convolutional Networks (link).

    • CNN example (link).

    • Data Augmentation (link).

    • Leaderboard (link).

    • Deadline: November 09

Week 6 - November 02 - Holiday or Recess

Week 7 - November 09

Lecturer:

  • Heitor.

Content:

  • (SYNC) Transfer Learning (slides, video).

  • (SYNC) Final Project partial discussion.

Assignments:

  • (SYNC) Transfer Learning (link).

  • Deadline: November 25

Week 8 - November 16

Lecturer:

  • Andrei.

Content:

  • (ASYNC) Sequence Modeling, Recurrent and Recursive Networks (slides, video).

Assignments:

  • (SYNC) Sequence Modeling, Recurrent and Recursive Networks (link).

    • Example (link).

    • Deadline: December 03

Week 9 - November 23

Lecturer:

  • Andrei.

Content:

Assignments:

  • (SYNC) Natural Language Processing (link).

    • Deadline: December 10

Week 10 - November 30

Lecturer:

  • Anderson e Matheus.

Content:

Assignments:

  • (SYNC) Segmentation (link).

    • Deadline: December 17

Week 11 - December 7

Lecturer:

  • Heitor e André.

Content:

  • (SYNC) Final Project and Assignment Discussions.

  • (ASYNC) Other Recent Topics (watch them if necessary for your work):

Assignments:

  • No assignments.

Week 12 - December 14

Lecturer:

  • Heitor e André.

Content:

  • (SYNC) Final Project and Assignment Discussions.

Final Project and Grades

Content:

  • Assignment Grades and Final Project Proposals (link).

  • Dates and rules:

    • Final Report Deadline: February 07.

      • Must contain: Introduction (detailing the addressed problem); Methodology; Results (with comparisons, if possible); Conclusions.

      • Suggested template: IEEE.

    • Presentation Deadline: 14-18 Feb.

      • 15 minutes for each work + 5 minutes for questions.