Practical Deep Learning for Coders

What the course covers

Summarised by TrustMyCourse from the material reviewed. Wording is our own, not the provider’s.

1. 01

   Getting started

   Training a usable image model in the first session, before any of the underlying theory.

2. 02

   Deployment

   Putting a trained model somewhere other people can use it, early rather than at the end.

3. 03

   Neural network foundations

   What the training loop is actually doing, revisited once a model has already been built.

4. 04

   Natural language processing

   Applying the same approach to text, including working with pretrained transformer models.

5. 05

   Building a model from scratch

   Reconstructing the pieces the library normally provides, to show what it is abstracting.

6. 06

   Random forests and tabular data

   Tree-based methods for structured data, where deep learning is not automatically the right tool.

7. 07

   Collaborative filtering

   Recommendation problems and the embedding ideas behind them.

8. 08

   Convolutions and CNNs

   The architecture behind the vision results from the opening lessons.

9. 09

   Data ethics

   A bonus unit on the consequences of deploying models, included alongside the technical lessons.