7 Deep learning 1: Backpropagation for tensors, Convolutional Neural Networks (MLVU2019) @riskone1

Uploaded February 2019 | Updated September 2026, 4 days ago

slides: mlvu.github.io/lectures/41.DeepLearning1.annotated.pdf
course materials: mlvu.github.io

This lecture builds on the explanation of backpropagation in the previous lecture: youtu.be/g2lziWxf_9Q

We are finally ready to discuss Deep Learning. We take the basic concepts from the last lecture on Neural Networks, and we show how to build them into a proper deep learning framework: an environment that lets us build networks that are many layers deep, with up to billions of parameters, and still train the whole thing with gradient descent.

The explanation broadly follows the structure of PyTorch, but most concepts apply to Tensorflow and other deep learning frameworks too.

7 Deep learning 1: Backpropagation for tensors, Convolutional Neural Networks (MLVU2019)

8 Probability 2: Density estimation. Maximum likelihood, Normal Distributions,

13 Reinforcement Learning: Policy Gradients, Q Learning, AlphaGo (MLVU2018)

04 Methodology 2: Data cleaning, Principal Component Analysis (MLVU2018)

MLVU 1.3 Other abstract tasks: regression, clustering, density estimation

MLVU 9.2: Generative adversarial networks (GANs)

1 Introduction to Machine Learning (MLVU2020)

8 Probability 2: Maximum Likelihood, Gaussian Mixture Models and Expectation Maximization (MLVU2019)