From the course: TensorFlow: Practical Skills in Constructing, Training, and Optimizing Models

Unlock this course with a free trial

Join today to access over 25,300 courses taught by industry experts.

Implement a neural network

Implement a neural network - TensorFlow Tutorial

From the course: TensorFlow: Practical Skills in Constructing, Training, and Optimizing Models

Start my 1-month free trial Buy for my team

Implement a neural network

“

- [Instructor] Neural networks are some of the highest-performing models in machine learning today, and you can build them with TensorFlow. In this lesson, I'll walk you through how to build a simple neural network in TensorFlow. Open up 04_01_Implement.ipynb. The first cell imports the familiar tools, as well as TensorFlow's radon levels dataset. The next cell prepares that data for modeling by loading the dataset, isolating a feature and target from the dataset, and making sure they have the right shape for the model. It may take a minute or more to execute the code in this cell. Then the next cell builds my model. I use the sequential model type to start my network. This model type is appropriate whenever you need a model with single inputs and single outputs, and it's also very user-friendly. Next, I add three dense layers to flesh out the model structure. By dense, what I mean is that every node in the layer is connected to every node in the next layer. This is how you can create…

Contents