Skip to content

Neural Networks and Deep Learning

40 hoursIntermediate7 modules

4 of 7 modules with a published syllabus

0 of 7 modules recorded

Syllabus

Goes from visual intuition to implementation: gradients and applied linear algebra, a neural network written from scratch with no framework, the same network rewritten in PyTorch with autograd and GPU, and the discipline of evaluation and experimentation. The two area modules add CNNs — shared with Computer Vision — and regularisation/normalisation.

Modules of this programme

The same objects as the full curriculum. A module opened from any programme that contains it is the same module, with the same syllabus and the same status.

  • DL-101Syllabus published

    Visual intuition for neural networks

    Detailed syllabus not written yet

  • CORE-201Syllabus published

    Essential mathematics II

    Content
    Gradients, optimisation, applied linear algebra
    Hands-on
    Gradient descent from scratch in NumPy

    estimated price0 of 1 module published

  • CORE-202Syllabus published

    Neural networks from scratch

    Content
    Perceptron, backpropagation, activation functions
    Hands-on
    Implement a neural network with no framework

    estimated price0 of 1 module published

  • CORE-203Syllabus published

    PyTorch in practice

    Content
    Tensors, autograd, training loops, GPU
    Hands-on
    Rewrite the CORE-202 network in PyTorch

    estimated price0 of 1 module published

  • CORE-204Syllabus published

    Evaluation and experimentation

    Content
    Overfitting, cross-validation, experiment tracking (W&B/MLflow)
    Hands-on
    Run and compare 10 experiments

    estimated price0 of 1 module published

  • DL-201Syllabus published

    CNNs (shared with Computer Vision)

    Detailed syllabus not written yet

  • DL-203Syllabus published

    Regularisation and normalisation

    Detailed syllabus not written yet

Not everything fits into 40 hours. State of the art, distributed training, interpretability, edge deployment and the five integrative projects exist only inside the complete tracks: Tracks