Machine Learning Platforms

May 7, 2018

 

 

There are many methods for doing automated Machine Learning. Early methods modeled biological or statistical building blocks in low level procedural code. The programming methodologies have evolved over the years to have libraries of the building blocks built and packaged for convenience. Even at a higher level, whole platforms have been packaged and provided in such a way that almost drag and drop methods can be used, with parameter choice made at different levels of automation.

 

 

Languages

 

Python

R

SQL

SAS

others (C/C++, java, etc.)

 

Libraries

 

Tensorflow (Google)

Keras

CoreML (Apple)

  https://machinelearning.apple.com/

  https://developer.apple.com/machine-learning

OpenCV

Pandas

Scikit-learn

MxNet

 

 

Big Data Platforms

 

CUDA (Program on NVIDIA GPUs)

 

Even Easier Introduction CUDA

GPU Accelerated Computing with Python

 

Hadoop (Cluster management using MapReduce algorithm)

Spark/Skala (New cluster management)

 

 

Cloud Services

    (most available on-prem as well)

 

Google's AutoML

 

DataScience.com

  Oracle Buys DataScience.com


DataRobot

 

Ersatz

 

Alteryx

 

C3iot.ai

 

H2O.ai

 

IBM Watson

 

Please reload

Our Recent Posts

June 11, 2018

Please reload

Archive

Please reload

Tags

Please reload

©2020 by Deep Side of Learning. Proudly created with Wix.com