A Brief Introduction to TensorFlow for Machine Learning

Original article was published by Komal Saini on Artificial Intelligence on Medium

A Brief Introduction to TensorFlow for Machine Learning

A free, open-source framework popularly used for machine learning.


TensorFlow is an open-source software library for machine learning, originally developed by the Google Brain Team.

How does TensorFlow work?

TensorFlow allows programmers to build data-flow graphs, which describe how data moves through a graph database. Each node in the graph is representative of a mathematical operation, and each connection between nodes is a multidimensional data array (aka a tensor).

(a) Example data-flow graph, (b) Path extracted from the graph (Source)

TensorFlow utilizes the Python programming language to provide this; nodes and tensors are Python objects, and TensorFlow applications are Python applications; however, the mathematical operations are written as high-performance C++ binaries.

TensorFlow can train and run deep neural networks for image recognition, natural language processing, and classification, to name a few.

Who uses TensorFlow?

Thousands of developers and over 300 companies —including Uber, Twitter, and Airbnb — use TensorFlow in their tech stack.

  • Airbnb improves guest experience by using TensorFlow to classify images and to detect objects at scale.
  • Twitter built their “Ranked Timeline” feature using TensorFlow to bring the most important tweets to the attention of each user, even when following thousands of others.
  • GE Healthcare used TensorFlow to train neural networks to identify specific anatomical structures during MRIs, improving their speed and reliability.

Why is TensorFlow such a popular machine learning framework?

One of the most significant benefits of TensorFlow is abstraction; developers primarily focus on the logic of the application, and TensorFlow deals with the details surrounding the implementation of algorithms and other inconveniences. The TensorBoard visualization suite also allows developers to view graphs through an interactive, web-based dashboard.


How can I gain practical experience as a beginner?

If you’re interested in machine learning, you must first build a strong understanding of math and stats, machine learning theory, and programming. Like most other skills, becoming an expert requires lots of hands-on experience to put your knowledge to the test.

Google Colab is a great way to get some practice; it is a free Jupyter notebook environment that runs in the cloud.

Here’s a basic classification guide that you can run in Google Colab.

Here’s a great resource to learn the basics of machine learning with TensorFlow. Although software development experience in Python is required to understand the learning materials, it is a beginner-friendly guide for those who are new to machine learning, but have a background in development and computer science.

To learn more, check these out: