Skip to content

2026.2

Instructor
Humberto Sandmann humbertors@insper.edu.br

Schedule
Lecture - -h00 -h00
Lecture - -h00 -h00
Office Hours - -h00 -h30

Final Grade

\[ \text{Final} = \left\{\begin{array}{lll} \text{Individual} \geq 5 \bigwedge \text{Team} \geq 5 & \implies & \displaystyle \frac{ \text{Individual} + \text{Team} } {2} \\ \\ \text{Otherwise} & \implies & \min\left(\text{Individual}, \text{Team}\right) \end{array}\right. \]

Syllabus 2026.2

Module 1
Foundations
  • Concepts & AI
  • Data
  • Preprocessing
  • Neural Networks
  • Perceptron
  • MLP
  • Optimization
  • Regularization
  • Metrics
Module 2
Deep Architectures
  • DL Layers
  • CNNs
  • Attention NEW
  • Transformers NEW
  • Transfer Learning NEW
  • LLMs NEW
Module 3
Generative Models
  • Overview
  • VAE
  • GAN
  • CLIP
  • Stable Diffusion
  • Flow Matching
  • Diffusion Transformers NEW
  • AR Generation NEW

Course Description

This course provides a comprehensive introduction to Artificial Neural Networks and Deep Learning, using modern frameworks (primarily PyTorch). Topics span mathematical foundations, core architectures (MLPs, CNNs, Transformers), attention mechanisms, generative models (GANs, VAEs, Diffusion, Flow Matching), Diffusion Transformers, and Large Language Models. Equal emphasis is placed on theoretical understanding and practical application.

Learning Objectives

By the end of this course, students will be able to:

  1. Understand Fundamentals: explain gradient descent, backpropagation, activation functions, and regularization.
  2. Master Key Architectures: describe and motivate MLPs, CNNs, Transformers, and LLMs.
  3. Implement with PyTorch: train and debug deep learning models.
  4. Evaluate Performance: apply appropriate metrics and regularization/optimization techniques.
  5. Work with Generative Models: understand and apply GANs, VAEs, Diffusion, and Flow Matching.
  6. Apply Transfer Learning: fine-tune pre-trained models using PEFT techniques (LoRA, QLoRA).
  7. Understand LLMs: grasp architecture, training (RLHF, DPO), and applications of Large Language Models.
  8. Critically Evaluate Research: read and assess current deep learning papers.

Bibliography

Core:

  1. Fleuret, F. (2023). The Little Book of Deep Learning.
  2. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press.

Supplementary:

  1. Nielsen, M. A. (2019). Neural Networks and Deep Learning.
  2. Zhang, A. et al. (2024). Dive into Deep Learning.
  3. Vaswani, A. et al. (2017). Attention Is All You Need.
  4. Brown, T. et al. (2020). Language Models are Few-Shot Learners.