Fundamentals of Neural Networks - Lab 5 - Building Deeper and Wider Model

This lab session covers building a wider and deeper neural network architecture using the functional API. It contrasts the functional API with the sequential API, highlighting the flexibility of the former in creating complex architectures. The session details the definition of input paths, hidden layers, and the use of the Raylu activation function. It explains the convergence of paths and the handling of multiple outputs, emphasizing the importance of matching dimensions in the training set. The session concludes with implementing the architecture using class objects in Python for cleaner code.

What are the two outputs produced by the model, and how are they defined?

What considerations must be made regarding the training set for the model?

Summarize the overall architecture discussed in the lab session.

How does the class object help in organizing the code for the neural network?