Credit: Google News
The “Intro to TensorFlow for Deep Learning” course is designed to be accessible to developers without a math background.
Google and online learning hub Udacity have launched a free course designed to make it simpler for software developers to grasp the fundamentals of machine learning.
The “Intro to TensorFlow for Deep Learning” course is designed to be more accessible to developers than previous machine-learning courses offered by Udacity.
“Our goal is to get you building state-of-the-art AI applications as fast as possible, without requiring a background in math,” says Mat Leonard, head of the School of AI at Udacity.
“If you can code, you can build AI with TensorFlow. You’ll get hands-on experience using TensorFlow to implement state-of-the-art image classifiers and other deep learning models. You’ll also learn how to deploy your models to various environments including browsers, phones, and the cloud.”
SEE: Hiring kit: Python developer (Tech Pro Research)
The course will teach students how to use the latest version of Google’s TensorFlow machine-learning framework, version 2.0 alpha, will walk them through core machine learning concepts, and cover how to build and train a neural network, the brain-inspired mathematical models that underpin deep learning.
The two-month online course will start with a simple example of using a neural network to convert temperature from Celsius to Fahrenheit, progressing to cover training a deep neural network to recognize items of clothing, while also exploring how Convolutional Neural Networks work and covering topics such as data augmentation and transfer learning. It’ll also detail how to deploy trained deep-learning models on browsers, Android, iOS, and single-board computers like the Raspberry Pi.
Teaching takes places via exercises and Colab notebooks written by the TensorFlow team, which will illustrate some of the most common applications of neural networks.
Udacity’s Leonard attributes much of the simplicity of the course to TensorFlow 2.0, which revamps the framework to make it simpler to use.
TensorFlow 2.0 introduces eager execution by default, which allows tasks to be accomplished using fewer lines of code and makes debugging easier — with the aim of making TensorFlow as simple to use as something like the open-source machine learning library Pytorch.
It also cleans up deprecated APIs, in favor of making TensorFlow accessible via the Keras API, which is prized for allowing developers to rapidly prototype, test and extend machine-learning models.
Keras offers a high-level API that streamlines the process of building a neural network by splitting the building blocks of the networks into reusable classes and abstracting away some of the details of creating the network.
The free course is part of Udacity’s School of AI, a set of courses and Nanodegree programs designed for software developers, which also covers topics such as linear algebra and calculus, foundational machine-learning models, and state-of-the-art deep learning.
Machine learning is obviously a massive topic, so it’s likely this latest two-month course will likely only be able to offer a grounding in the fundamentals.
Most of the major technology companies offer free machine-learning courses that educate developers about the basics of the field, likely in a bid to establish their own software frameworks as the standard in the industry.
You can read more about the free machine-learning courses backed by the likes of Microsoft, Google and Amazon and how they compare in this TechRepublic overview.
If you’re interested in which machine-learning related libraries are available for your programming language of choice, then check out TechRepublic’s round-up of the most highly rated machine learning libraries on GitHub.
Credit: Google News