Why is Python so popular in machine learning?

Python is like the swiss army knife of machine learning.

It is not the fastest. Scala, Julia is faster and most of the heavy lifting in Python is actually done by C or Fortran libraries backend. Neither is it the easiest to learn. R is easier for beginners. Rather, it is a general language that does a little of everything at a good enough complexity-performance tradeoff with a full suite of tools for productionizing machine learning.

Tutorials on machine learning with python

1. dive into machine learning
2. ?hat | Data Science in Python

A sample process of machine learning deployment with python:

1. Build prototypes on laptops with scikit-learn: machine learning in Python,NumPy, Python Data Analysis Library in an interactive Project Jupyternotebook with python plotting - Matplotlib visualization. Examples - List of interesting notebooks, more interesting notebooks, Probabilistic-Programming-and-Bayesian-Methods-for-Hackers.
2. Train the full model on Hadoop cluster with PySpark and MLlib | Apache Spark.Example - train a recommender on spark
3. Setup an API server with CherryPy - Minimalist Python Web Framework, build aFlask (A Python Microframework) app around the model and serve the model out as a REST api to other applications. Examples: Deploy recommender | Dockerized Data Science

Furthermore, advanced Python modules extend the capabilities:

Interactive plot extensions

• Bokeh
• plotly

Deep learning

• Deep learning framework by Google: https://www.tensorflow.org/
• TensorFlow + scikit-lean = skflow
• Pylearn2 + Theano

Natural Language Processing (NLP) / text analytics / parts of speech tagging

• Natural Language Toolkit
• https://spacy.io/
• TextBlob: Simplified Text Processing
• Pineapple: proycon/pynlpl
• gensim: topic modelling for humans

For more data related posts, visit my quora page.?