The Python Machine Learning Ecosystem

Abstract

In the first chapter we explored the absolute basics of Machine Learning and looked at some of the algorithms that we can use. Machine Learning is a very popular and relevant topic in the world of technology today. Hence we have a very diverse and varied support for Machine Learning in terms of programming languages and frameworks. There are Machine Learning libraries for almost all popular languages including C++, R, Julia, Scala, Python, etc. In this chapter we try to justify why Python is an apt language for Machine Learning. Once we have argued our selection logically, we give you a brief introduction to the Python Machine Learning (ML) ecosystem. This Python ML ecosystem is a collection of libraries that enable the developers to extract and transform data, perform data wrangling operations, apply existing robust Machine Learning algorithms and also develop custom algorithms easily. These libraries include numpy, scipy, pandas, scikit-learn, statsmodels, tensorflow, keras, and so on. We cover several of these libraries in a nutshell so that the user will have some familiarity with the basics of each of these libraries. These will be used extensively in the later chapters of the book. An important thing to keep in mind here is that the purpose of this chapter is to acquaint you with the diverse set of frameworks and libraries in the Python ML ecosystem to get an idea of what can be leveraged to solve Machine Learning problems. We enrich the content with useful links that you can refer to for extensive documentation and tutorials. We assume some basic proficiency with Python and programming in general. All the code snippets and examples used in this chapter is available in the GitHub repository for this book at https://github.com/dipanjanS/practical-machine-learning-with-python under the directory/folder for Chapter 2. You can refer to the Python file named python_ml_ecosystem.py for all the examples used in this chapter and try the examples as you read this chapter or you can even refer to the jupyter notebook named The Python Machine Learning Ecosystem.ipynb for a more interactive experience.