Deep Learning

Status: 🚧 Coming soon — chapters are being written.

Deep learning uses neural networks with many layers to learn representations from data. It's behind modern computer vision, NLP, speech, and the LLMs you've been studying in the GEN AI section.

What this section will cover

• Neural network fundamentals — neurons, layers, activations, backprop
• Frameworks — PyTorch (primary), TensorFlow/Keras for comparison
• Optimization — SGD, momentum, Adam, learning-rate schedules
• Regularization — dropout, batch norm, weight decay, data augmentation
• Architectures —
  • CNNs for vision
  • RNNs / LSTMs / GRUs for sequences
  • Transformers (the architecture behind LLMs)
• Transfer learning, fine-tuning, parameter-efficient fine-tuning (LoRA / QLoRA)
• Training tricks: mixed precision, gradient accumulation, distributed training

Currently available — related material

Deep-learning-adjacent context lives in these existing sections:

• Machine Learning → Linear Regression (gradient descent foundation)
• Machine Learning → Linear Regression — types of gradient descent
• LangChain → Models

A dedicated deep-learning track (PyTorch + transformers) is next.