The world’s leading publication for data science, AI, and ML professionals.

3 Deep Learning Projects Using Keras That You Can Complete Today

Step-by-step projects that you can add to your resume

Photo by Sebastian Pantosin on Unsplash
Photo by Sebastian Pantosin on Unsplash

Keras. If you’re entering the machine learning field or have taken on the challenge of learning how to program in Python, you might have heard of this library, and its industry standard companion Tensorflow.

Keras is used for thousands of professional applications all across the globe, and technologies that take advantage of the Keras API are probably embedded in various apps in your smartphone and computer.

So, what is Keras, and why do I believe that it is an absolutely essential API and library to learn, whether you are a beginner coder or seasoned Python pro? Let’s find out.


What exactly is Keras?

Keras is a machine learning API and library built for integration with Python programs. It contains various modules and shortcuts for building and testing your own machine learning models, at virtually any level of intensity. It can be used for very simple, surface learning tasks, or heavy Deep Learning tasks where complex analysis is required.

Keras is predominantly built for neural networks, which makes it extremely versatile for many tasks. For example, binary classification and image classification very often depend on the Keras library. With Keras, there are so many shortcuts for creating neural network classes and configuring them, and you can get up and running with a machine learning program in about half an hour for most tasks.

With about four hundred thousand regular users and a huge network of endorsers, Keras is favored thanks to its simplicity and nature, which allows both beginners and experts to create handy learning models. Such programs can easily be deployed for real life, industrial practices, and over the past few years we have seen various Keras models utilized for medical predictions. Scroll down to the examples section to find out more innovative ways that the Keras library has been used for real world challenges!

But Keras isn’t simply a standalone Machine Learning library. It is an API that can work in harmony with software like Tensorflow – Google’s own machine learning kits. In some ways, Keras can be described as a lego-brick library. It’s ridiculously simple to add and connect neural network layers together using single lines of code, and performing train and test routines is hugely simplified thanks to prebuilt functions.


Keras Projects that You Can Complete Today

The Keras machine learning library is not just limited to amateur projects. It has been deployed hundreds of times in a massive range of real life applications, helping app developers improve their software, medical practices make better diagnoses, improving traffic systems, and much much more. In this section, we explore several outstanding programs built with the Keras library that show just how much you can take advantage of its pre-built features and push it to its limits!

Heart Disease Classification

Link to guide.

Photo by jesse orrico on Unsplash
Photo by jesse orrico on Unsplash

When you have a large enough dataset, you can predict virtually anything you want! In the medical field, Keras models have been used to make more accurate diagnoses based on patients’ conditions and specifics. Heart disease classification is one common project that has proven itself very useful and applicable when using the Keras library.

This excellent example shows how a developer in Tel Aviv, Israel created a machine learning model to predict the percentage chance that a patient had a heart disease based on just fifteen data categories. Impressive, right?

Rock Paper Scissors

Link to guide.

Photo by Marcus Wallis on Unsplash
Photo by Marcus Wallis on Unsplash

On a more trivial level, we have seen developers create machine learning models for simple childhood games, such as rock, paper, scissors. In this example, a convolutional neural network was used to classify images between the three categories of rock, paper and scissors to see what hand the competitor was playing. The fact that we have amateurs doing this kind of Programming for fun is a symbol of how far we’ve come since those first computer chips in the seventies.

Face Mask Detection

Link to guide.

Photo by Mika Baumeister on Unsplash
Photo by Mika Baumeister on Unsplash

It’s great to see machine learning models built and deployed for relevant and helpful applications – in this case, detecting whether the photographed person is wearing a face mask or not. This is another image classification model, and believe it or not, there are plenty of tutorials online for you to learn how to replicate this yourself. Similarly, programs like this have been used for facial recognition for many years, and the power of the Keras library really begins to show when it is paired up with image processing tools like OpenCV.


Thanks for Reading!

By reading this article, you should have a fundamental understanding of what Keras is, as well as how it can be used to create awesome projects that you can create with relative ease! I hope this inspires you to learn and create your own projects that you can add to your resume and GitHub.

As always, I wish you the best in you endeavors.

Not sure what to read next? I’ve picked another article for you:

A Complete 52 Week Curriculum to Become a Data Scientist in 2021

Terence Shin


Related Articles