Intro to deep learning
1 like282 views
The document introduces deep learning and provides examples of its applications. It discusses using a neural network to recognize handwritten digits, how cost functions are used to evaluate accuracy, and how gradient descent helps optimize the network. Potential uses of deep learning mentioned include generating text, images, music, code and designs, and personalizing experiences based on user behavior and expressions. The document also lists several datasets and playgrounds for experimenting with deep learning, as well as resources for using it in frontend development.
Intro to deep learning
- 1. Intro to Deep Learning Ruxandra Burtica ruxandra.burtica@gmail.com
- 3. What does learning look like?
- 5. Recognize handwritten digits (MNIST)
- 11. XW = Z X W Z
- 14. Cost function Small when the network classi鍖es correctly the digit Large otherwise Compute average cost for all training examples ==> that is our networks cost
- 15. Gradient descent
- 17. How it applies to web
- 18. Generate Text Pictures Music Code Designs
- 19. Generate Text Pictures Music Code Designs
- 20. Generate Text Pictures Music Code Designs
- 21. Generate Text Pictures Music Code Designs
- 22. Generate
- 23. Personalize Based on user behavior (previous interactions) Based on user expression (real-time )
- 24. Personalize Based on user behavior (previous interactions) Based on user expression (real-time )
- 25. Personalize Based on user behavior (previous interactions) Based on user expression (real-time )
- 26. Datasets
- 30. Playgrounds
- 33. deeplearnjs.org
- 34. Resources
- 35. https://deeplearnjs.org https://transcranial.github.io/keras-js http://ml4a.github.io/demos/forward_pass_mnist https://blog.鍖oydhub.com/turning-design-mockups-into- code-with-deep-learning https://blog.insightdatascience.com/automated-front- end-development-using-deep-learning-3169dd086e82 https://github.com/Morphl-Project