cyber2a / cyber2a-course

Online materials for the Cyber2A course on AI for Arctic research
https://cyber2a.github.io/cyber2a-course/
Apache License 2.0
0 stars 1 forks source link

Lesson - Intro to Neural Networks #4

Closed carmengg closed 7 months ago

carmengg commented 12 months ago

Goal

Gain a high-level understanding of how a NN works. Understand main components of NN input data, layers, weights, targets, loss function, optimizers, train and test sets, ...

Breakdown

  1. Introduction and Overview
    • Brief history of NN
    • Importance and applications in today's world
    • Relevance to Arctic science
  2. Basic concepts and terminology
    • Neurons and layers: Input, Hidden, Output
    • Weights and biases
    • Activation functions: Sigmoid, ReLU, etc.
  3. How Neural Networks Learn
    • Forward propagation: How input becomes output
    • Cost function: Measuring how "wrong" the network is
    • Backpropagation: Adjusting weights and biases
    • Gradient descent and learning rate
  4. Types of Neural Networks
    • MLP
    • CNN
    • RNN/LSTM
    • Transformer
  5. Training, Validation, and Testing
    • Splitting data: Why and how
    • Overfitting and underfitting: Concepts and solutions
    • Regularization techniques
  6. Real-world Example and Demonstration
    • Introduce frameworks (PyTorch, Tensorflow)
    • simple/interesting examples
    • Visualization