APPENDIX A
References
In this appendix, I provide references to books, papers, and online articles that cover many topics in detail that I have mentioned in the book. I list the references by each chapter and include references that are mostly free and include code samples that you can readily use. Thanks to the absolutely amazing Deep Learning community, there are many such resources readily available to you. As long as you understand the basic concepts around ML and DL, you should be able to follow these references and the code.
Chapter 1: Big Data and Artificial Intelligence
- Dr. Andrew Ng from Stanford is one of the foremost researchers in Machine Learning and AI. I highly recommend his videos and news items that define the state of AI.
https://www.deeplearning.ai/the-state-of-artificial-intelligence-andrew-ng-at-mit-emtech-2017/ - General Electric has been leading the Big Data revolution in the industrial space with the rise of the industrial Internet. Here is an excellent whitepaper on the prospect of industrial IoT from GE.
- The industrial IoT revolution is known as Industry 4.0 especially in Europe. Here is a nice article by Bernard Marr explaining this.
- This is a nice article showing how Amazon transformed itself around Artificial Intelligence. It includes a great example of driving a platform vision to improve several products at once.
https://www.wired.com/story/amazon-artificial-intelligence-flywheel - There have been cool developments in the news about modern AI generating images of paintings and fake celebrities.
AI-generated art sold for $432,500!
AI generates fake celebrity photos.
https://www.theverge.com/2017/10/30/16569402/ai-generate-fake-faces-celebs-nvidia-gan - Finally, I recommend visiting the AI research pages of these top companies. They often have amazing content. Here are some pages I often visit.
- Google
https://ai.google/ - Facebook
https://onnx.ai - NVIDIA
https://www.nvidia.com/en-gb/deep-learning-ai/ - Intel
https://software.intel.com/en-us/ai-academy - IBM Watson
https://www.ibm.com/watson/ - Salesforce Einstein
https://www.salesforce.com/products/einstein/overview/ - H2O:
https://www.h2o.ai/
- Google
Chapter 2: Machine Learning
- Whenever someone asks me for a good starting place to learn ML and DL, my first reference is always the video course by Dr. Andrew Ng. This is universally accepted as the foremost resource for learning ML. It explains the different algorithms extremely well, giving you a lot of details of the basic concepts. You can join a certificate course at the following sites. The certification costs, but you can see the course videos for free.
https://www.coursera.org/learn/machine-learningSome parts of the course material are also available for free on YouTube.
https://www.youtube.com/user/StanfordUniversity - Google offers an online free crash course in ML that is quite good.
https://developers.google.com/machine-learning/crash-course/ml-intro - One of my personal favorites is the podcast on basics of ML by Tyler Renelle. I found myself hooked to this one due to Tyler's simplistic way of explaining several concepts. Highly recommended.
- Here are a couple more recommended podcasts.
https://www.thetalkingmachines.com/ - Another great site with awesome ML articles and sample code is Analytics Vidhya. This was founded by Kunal Jain and provides great tutorials. Here are some I found extremely helpful:
https://www.analyticsvidhya.com/blog/2017/09/common-machine-learning-algorithms/https://www.analyticsvidhya.com/blog/2016/01/complete-tutorial-learn-data-science-python-scratch-2/https://www.analyticsvidhya.com/blog/2018/03/comprehensive-collection-deep-learning-datasets/ - Kaggle is an amazing resource for ML practitioners. It has competitions hosted by companies where they provide good datasets that you can analyze. You compete with data scientists around the world to build models with highest accuracy, scored against a predefined result. It's a great way to explore your data science skills and play with real-world data, and many of the competitions involve cash prizes. Highly recommended.
Chapter 3: Handling Unstructured Data
- Computer vision details and tutorials are available at the OpenCV tutorials website. I prefer Python, but they have tutorials for C++ and Java also.
https://docs.opencv.org/3.0-beta/doc/py_tutorials/py_tutorials.html - Another fantastic resource for computer vision tutorials with some extremely well-written code is this site by Adrian Rosebrock. I particularly appreciate the sample code Adrian provides because it is highly generalized and easy to reuse.
https://www.pyimagesearch.com/Adrian also conducts a crash course in computer vision that is pretty good.
- For core ML algorithms, the library we use (Scikit-Learn) also provides some very good tutorials.
- For Natural Language Processing (NLP), there is an online tutorial by the NLP Toolkit that is pretty comprehensive.
- Also for NLP, I found this particular article straightforward and helpful.
https://dzone.com/articles/nlp-tutorial-using-python-nltk-simple-examples - Here is a nice primer on using Jupyter Notebooks.
Chapter 4: Deep Learning Using Keras
- For Deep Learning, I highly recommend the Andrew Ng video course mentioned previously. The concepts are explained beautifully and it's the best way to get started.
- Keras has been officially recognized as the premier frontend library to TensorFlow. The TensorFlow website now has some very good code examples that explain building deep networks using Keras.
- For Deep Learning with Keras, I recommend the book Deep Learning with Python by none other than the founder of Keras, François Chollet.
- Here's a list of useful Keras resources available at François Chollet's GitHub page:
Chapter 5: Advanced Deep Learning
- Another great resource on Deep Learning is the book TensorFlow for Deep Learning by Reza Zadeh and Bharath Ramsundar. I particularly recommend the “Reinforcement Learning” chapter in this book.
- The website KDNuggets.com has a set of very useful Deep Learning articles using Keras.
https://www.kdnuggets.com/2017/10/seven-steps-deep-learning-keras.html
Chapter 6: Cutting-Edge Deep Learning Projects
- The technical paper, “A Neural Algorithm of Artistic Style,” by Leon A. Gatys, Alexander S. Ecker, and Matthias Bethge, is a great resource.
- Check out the neural style transfer post with sample code by Raymond Yuan, from the TensorFlow team.
- The technical paper, “Generative Adversarial Networks,” by Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio, is also a good resource.
- Check out the article on generative adversarial networks with sample code from my favorite website, Analytics Vidhya.
https://www.analyticsvidhya.com/blog/2017/06/introductory-generative-adversarial-networks-gans/ - Check out this fantastic article by Dr. David Ellison on fraud detection using auto encoders in Keras.
https://www.datascience.com/blog/fraud-detection-with-tensorflow
Chapter 7: AI in the Modern Software World
- The Kubernetes website provides an excellent and interactive tutorial on setting up a cluster, including the basic commands. This is a great way to get familiar with the interface without doing the install on your machine.
- Another great website that provides great interactive tutorials is Katacoda. You get the same interface that you see in a production Kubernetes install, but can safely try all the commands. It's an amazing way to teach a technology.
Chapter 8: Deploying AI Models as Microservices
- Check out this excellent overview tutorial on microservices architecture by Martin Fowler and James Lewis. There's no code here, but it includes an amazing explanation of core concepts defining the architecture.
- Here's my GitHub repository, with an open source toolkit for quickly converting an image processing Keras model to a microservice hosted on Flask.
Chapter 9: Machine Learning Development Lifecycle
- Here is a nice whitepaper by Mesosphere on building an end-to-end platform for data science. It covers several of the data science concerns and how a platform can address them.
https://mesosphere.com/resources/building-data-science-platform/ - Check out “The 7 Steps of Machine Learning” video by the Google Cloud Platform team.
- Check out the blog post, “Data Scientists and Deploying Machine Learning into Production: Not a Great Match,” from Algorithmia.
Chapter 10: A Platform for Machine Learning
- Check out this Intel blog on training a TensorFlow model and deploying on Kubernetes.
- This is a nice article from the Google TensorFlow team called “Serving ML Quickly with TensorFlow Serving and Docker,” by Gautam Vasudevan and Abhijit Karmarkar.
https://medium.com/tensorflow/serving-ml-quickly-with-tensorflow-serving-and-docker-7df7094aa008 - Check out Katacoda's interactive tutorial on deploying Machine Learning workloads using Kubeflow and Kubernetes.
https://www.katacoda.com/kubeflow/scenarios/deploying-kubeflow