deep learning as a service what is it that users want
play

Deep Learning as a Service What is it that users want? Wolfgang zu - PowerPoint PPT Presentation

Deep Learning as a Service What is it that users want? Wolfgang zu Castell castell@helmholtz-muenchen.de Helmholtz Zentrum Mnchen DEEP-Hybrid-DataCloud has received funding from the European Unions Horizon 2020 research and innovation


  1. Deep Learning as a Service What is it that users want? Wolfgang zu Castell castell@helmholtz-muenchen.de Helmholtz Zentrum München DEEP-Hybrid-DataCloud has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 777435.

  2. Do we still need science? Chris Anderson, Wired 2008 acloud.eu Oct 10, 2018 2/14 https://deep-hybrid-dat atac

  3. Do we still need science? Chris Anderson, Wired 2008 acloud.eu Oct 10, 2018 3/14 https://deep-hybrid-dat atac

  4. Artificial neural networks Source: P . Shrivastava acloud.eu Oct 10, 2018 4/14 https://deep-hybrid-dat atac

  5. Clarifying terminology Deep Representation Machine Learning Learning Artificial Learning Intelligence acloud.eu Oct 10, 2018 5/14 https://deep-hybrid-dat atac

  6. Increase in number of neurons 1 Perceptron 4 Rumelhart 11 GPU accel. CNN 12 Deep Boltzmann 18 Multi-GPU CNN 20 GoogLeNet Source: I. Goodfellow acloud.eu Oct 10, 2018 6/14 https://deep-hybrid-dat atac

  7. Increase in size of data sets Source: I. Goodfellow acloud.eu Oct 10, 2018 7/14 https://deep-hybrid-dat atac

  8. How does deep learning work? functional mapping of pixels onto objects is very complicated deep learning is solving this issue by breaking up the problem into series of nested simple mappings Source: I. Goodfellow acloud.eu Oct 10, 2018 8/14 https://deep-hybrid-dat atac

  9. Machine learning cycle acloud.eu Oct 10, 2018 9/14 https://deep-hybrid-dat atac

  10. Who is the user? mac achine lear arning expertise technological expertise level of knowledge being domain expertise required depends on the specific use case acloud.eu Oct 10, 2018 10/14 https://deep-hybrid-dat atac

  11. Retinopathy use case Step 1: provide trained network such that others can use it for prediction Step 2: provide trained network for new classification task (transfer learning) Step 3: completely retrain network on new dataset Step 4: completely retrain network under data privacy constraints Source: Eulenberg et al. acloud.eu Oct 10, 2018 11/14 https://deep-hybrid-dat atac

  12. The ecosystem of tools & libraries ensorFlow: speech and image T recognition (Google Brain T eam) computer vision Keras as: Python NN library (Francois Challet, Google) PyT orch: DL library (Facebook KI) num. lin. alg. Caf affe: DL library (UC Berkeley) mxnet: scalable DL framework sci. comp. (Apache) plotting acloud.eu Oct 10, 2018 12/14 https://deep-hybrid-dat atac

  13. Architecture for deep learning acloud.eu Oct 10, 2018 13/14 https://deep-hybrid-dat atac

  14. Thank you Any Questions? https://deep-hybrid-datacloud.eu This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 777435.

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend