Deep Learning

Deep Learning

Deep Learning is taking hold as a popular machine learning modeling technique because of its real world applications especially with regards to image, signal and language datasets (e.g. medical diagnosis, self-driving cars, real-time language translation). This talk provides an overview of what deep learning is especially around recent applications.

2168aa4564112d3ba88869ca3cc994b3?s=128

Melanie Warrick

December 03, 2015
Tweet

Transcript

  1. Deep Learning Melanie Warrick Skymind @nyghtowl

  2. @nyghtowl Artificial Neural Nets Input Output Hidden Run until error

    stops improving = converge Loss Function
  3. @nyghtowl Deep Learning...

  4. @nyghtowl What Makes it Deep? Layers Feature Engineering

  5. None
  6. @nyghtowl Why DL Matters? - Feature Engineering - Language &

    Image Analysis - Unsupervised Modeling
  7. @nyghtowl Neural Nets...

  8. @nyghtowl Feed Forward Neural Net Who is it? Pixels Edges

    Object Parts Object Models Layer 2 Layer 3 Layer 4 Input
  9. @nyghtowl Convolutional Neural Net

  10. @nyghtowl Recurrent Neural Net

  11. @nyghtowl Restricted Boltzmann Machine

  12. @nyghtowl Multiple Models

  13. @nyghtowl Real World Value...

  14. None
  15. DUKE VINCENTIO: Well, your wit is in the care of

    side and that. Second Lord: They would be ruled after this chamber, and my fair nues begun out of the fact, to be conveyed, Whose noble souls I'll have the heart of the wars. Clown: Come, sir, I will make did behold your worship. VIOLA: I'll drink it.
  16. None
  17. @nyghtowl

  18. @nyghtowl

  19. @nyghtowl

  20. None
  21. @nyghtowl Last Points - deep neural nets - automation &

    personalization - unsupervised
  22. @nyghtowl • http://www.seattlepi.com/science/article/NASA-photos-as-hallucinated-by-Google-s-Deep-Dream-6654010.php#photo- 9009852 • http://www.wired.com/wp-content/uploads/2015/06/siri-ft.jpg • http://www.google.com/selfdrivingcar/images/home-where.jpg • http://i.telegraph.co.uk/multimedia/archive/02122/WILLIAM-SHAKESPEAR_2122089b.jpg

    • https://karpathy.github.io/2015/05/21/rnn-effectiveness/ • http://static.artdiscover.com/img/news/340_2_l.jpeg • http://www.wildml.com/2015/09/recurrent-neural-networks-tutorial-part-1-introduction-to-rnns/ • http://deeplearning.net/tutorial/lenet.html • http://web.media.mit.edu/~lieber/Teaching/Context/ • http://arxiv.org/pdf/1412.7755.pdf • https://www.kaggle.com/forums/f/15/kaggle-forum/t/10878/feature-representation-in-deep-learning • http://jaoying-google.blogspot.com/2012_12_01_archive.html • http://users.clas.ufl.edu/glue/longman/1/einstein.html • https://pbs.twimg.com/media/CJm9HmfVEAEXU0c.jpg:large • http://i.telegraph.co.uk/multimedia/archive/03370/broadway2_3370419k.jpg • https://qiita-image-store.s3.amazonaws.com/0/60969/41c3fa0d-418a-eafe-65cb-2fba2c74dd12.png References: Images
  23. @nyghtowl Deep Learning Melanie Warrick skymind.io (company) @nyghtowl

  24. @nyghtowl Special Thanks • Adam Gibson • Alex Black •

    Bryan Catanzaro • Charlie Tang • Chris Nicholson • Christian Fernandez • Cyprien Noel • Diogo Almeida • Erin O’Connell • Isabel Markl • Jason Morrison • Jeremy Dunck • Josh Patterson • Kelley Robinson • Lindsay Cade • Marissa Marquez • Mark Ettinger • Megan Speir • Meggie Mahnken • Paco Nathan • Phillip Culliton • Tarin Ziyaee • Tim Elser • Tiia Priilaid
  25. • Nature of Code: Neural Networks http://natureofcode.com/book/chapter-10-neural-networks/ • Neural Nets

    and Deep Learning http://neuralnetworksanddeeplearning.com/ • Neural Nets Demystified (Welch) http://www.welchlabs.com/blog/ • Theano Tutorial http://deeplearning.net/software/theano/tutorial/index.html#tutorial • “The State of Deep Learning in 2014” https://speakerdeck.com/ogrisel/the-state-of-deep- learning-in-2014 • “Hacker’s Guide to Neural Nets” https://karpathy.github.io/neuralnets/ • “Automated Image Capturing with ConvNets and Recurrent Nets” (Karpathy) http://cs.stanford. edu/people/karpathy/sfmltalk.pdf • Deeplearning.net • Wikipedia.org Where to go next... @nyghtowl