Deep-Learning-ICTS6361-2023

Deep Learning ICTS6361 2023

Course Outline

• Introduction To Neural Networks and Deep Learning
• Improving Deep Neural Networks, Hyperparameter Tuning, Regulaization and Optimization
• Structuring ML projects (end-to-end)
• Convolutional Neural Networks (CNN)
• Sequence Models (Recurrent Neural Networks RNNs) and Transformers

Evaluation:

• Selected Papers Presentation 30
• Project 30
• Final Exam 40

Textbooks

• Ian Goodfellow, Yoshua Bengio, and Aaron Courville. Deep Learning
• Understanding Deep Learning - Simon J.D. Prince

References

• DNN cheatsheet
• Deep Learning AI course
• Deep Learning Course at Stanford
• Deep Learning Course at MIT
• ML for everyone
• Math is Fun
• Log likelyhood
• A Comprehensive Introduction to Different Types of Convolutions in Deep Learning | by Kunlun Bai | Towards Data Science

Compare CNN architectures

• Different Types of CNN Architectures Explained: Examples, May 2023
• Illustrated: 10 CNN Architectures, 2019
• Best deep CNN architectures and their principles: from AlexNet to EfficientNet, 2021

Video Lectures

• Deep Learning Book
• Deep Neural Networks by Andrew NG (DeepLearning.ai)
  • DL.AI intro to DL
  • DL.AI Improving Deep Neural Networks
  • DL.AI Structuring Machine Learning Projects
  • DL.AI Convolutional Neural Networks
  • DL.AI Sequence Models
• Stanford Deep Learning Course
• MIT Deep Learning Course
• My viedo lectures

Pre-trained models repositories

• Tensorflow hub
• TensorFlow Models & datasets
• TensorFlow Models
• Keras Pre-Trained Models
• Hugging Face
• Pytorch models
• Pytorch Hub
• Model Zoo
• Papers with code

How to fine-tune (transfer learning)

• Fine Tune PyTorch Models
• Fine Tune Tensorflow Models
• Fine Tune Tensorflow Models with TF Hub
• Fine Tune HuggingFace Models
• Fine Tune HuggingFace Models with Keras