Blogs que discutem o assunto
- https://medium.com/datadriveninvestor/different-approaches-to-support-deep-learning-in-a-visual-programming-environment-c5c487ba4c7b
- https://towardsdatascience.com/making-deep-learning-user-friendly-possible-8fe3c1220f9
Ferramentas
- Ubuntu, Windows & Cloud: DLS Deep Leaning Studio da DeepCognition – http://deepcognition.ai/ – Deep Leaning Studio is a very interesting platform that has two operation modes: cloud and desktop. An exemple of building a DL model with DLS can be seen in this video: https://player.vimeo.com/video/198088116 Funciona com TF, Keras e o novo mxnet: https://deepcognition.ai/wp-content/uploads/2018/07/DLS-ProductOverview.pdf Importante: tem integration of pre-trained Keras models e Jupyter Notebooks, além de pre-configured environment for deep learning programmers.
- Ubuntu, Windows: Orange – https://orange.biolab.si/ É uma ferramenta visual de data mining que suporta DL e parece simples mas é também simples de instalar: em Linux instala com PIP – https://orange.biolab.si/download/ – https://orange.biolab.si/getting-started/ – https://www.youtube.com/channel/UClKKWBe2SCAEyv7ZNGhIe4g
- Cloud e Windows: SONY Neural Network Console https://dl.sony.com/– e (open source) https://github.com/sony/nnabla Usa as Sony Neural Network Libraries – https://nnabla.org/
- Cloud only: RapidMiner – https://rapidminer.com/ – RapidMiner has implemented an a-layer-as-a-component approach to integrate Keras into its platform. As shown in Figure 5, RapidMiner allows users to drag-and-drop a convolutional layer as a component, a pooling layer as a component, a dropout layer as a component, etc. At the time of writing, RapidMiner only support the sequential model in Keras – usa um cliente que tem de instalar, do mesmo jeito que as outras soluções cloud – nada é web-only – não consegui descobrir onde roda o cliente do RapidMiner , se é windows-only ou não
- Cloud only: IBM Neural Network Modeler – Watson Studio https://dataplatform.ibm.com/docs/content/analyze-data/ml_dlaas.html?audience=dr&context=refinery https://medium.com/ibm-watson/accelerate-your-deep-learning-experiments-with-ibms-neural-network-modeler-dd0c92fba814 Pareceu-me mais uma ferramenta para suportar serviços do Watson do que de fato uma coisa para ajudar gente que faz DL: https://www.youtube.com/watch?v=EDQ0AWcUBnE
- Windows-only: https://visualstudio.microsoft.com/downloads/ai-tools-vs/ – Visual Studio Tools for AI is an extension that supports deep learning frameworks including Microsoft Cognitive Toolkit (CNTK), Google TensorFlow, Theano, Keras, Caffe2 and more
IDE do DLS:
Links
Copyright © 2018 Aldo von Wangenheim/INCoD/Universidade Federal de Santa Catarina