ML Frameworks

I looked at various solutions AWS provided in their ML stack(more details at https://aws.amazon.com/amazon-ai/?nc2=h_l3_ai).  If you are an AWS shop and doesn't have much of ML/AI in-house expertise etc, it is a great place to start and learn and apply for simple projects.

However I have used following frameworks to build my last project.

Tensor Flow : https://www.tensorflow.org/ -  Most popular framework used for large-scale numeric computation. Born out of Google, used by various well-known companies and applications. As the name suggests, frameworks reduces complexity of numeric computation using data flow graphs.

Checkout TensorFlow Cheatsheet, this gives you great idea of 

https://github.com/kailashahirwar/cheatsheets-ai/blob/master/PDFs/Tensorflow.pdf?imm_mid=0f769c&cmp=em-data-na-na-newsltr_20171025

Apache Spark ML Library :  Very popular Open Source framework,  It is a module as part of Spark. Provides out-of box implementations and API for various ML algorithms(Classification, Regression, Clustering, Recommendation to name a few).

https://spark.apache.org/mllib/

You can read following article, which can be helpful knowing the landscape.

https://www.infoworld.com/article/2853707/robotics/11-open-source-tools-machine-learning.html#slide13

CoreNLP :  Standford publishes opensource CoreNLP library. Very useful to build basic NLP applications. A sample app could be parsing reviews from a website and take action based on that. 

https://stanfordnlp.github.io/CoreNLP/