Difficulty: beginner
Estimated Time: 15 minutes

In the previous scenarios, you've learned the central concepts of the Machine Learning process and created few layer for the neural network architecture. The network didn't perform well because we didn't apply any training yet.

In this scenario, we will do the walkthrough of the Deep Learning process. We will start from loading the data and building the layers. Then we'll use forward- and backpropagation in the training process.


You've completed Neural Network Training scenario.

You've learned how to accomplish the following tasks using Python numpy library:

  • Loading the data from the csv file
  • Building the fully connected layers for the Neural Network
  • Applying forward- and backpropagation
  • Training the network

Neural Network Training

Step 1 of 6

Neural Network Architecture

Let's build our first Deep Learning process. We will use the flower shaped dataset to accomplish this task. As we said, the data is not linearly separable. This means that algorithms that rely on the straight lines to separate classes won't be of much help.


When building a neural network for the case like this, it means we need to introduce hidden layers. The final architecture looks like the one in the image.

Network architecture

In the next steps we will load the data, build the network and train it to make future predictions.

This tab will not be visible to users and provides only information to help authors when creating content.

Creating Katacoda Scenarios

Thanks for creating Katacoda scenarios. This tab is designed to help you as an author have quick access the information you need when creating scenarios.

Here are some useful links to get you started.

Running Katacoda Workshops

If you are planning to use Katacoda for workshops, please contact [email protected] to arrange capacity.

Debugging Scenarios

Below is the response from any background scripts run or files uploaded. This stream can aid debugging scenarios.

If you still need assistance, please contact [email protected]