Summary
Throughout this chapter, we made a major stride to cover a respectable portion of the must-know elements of deep learning and neural networks. First, we investigated the roots of neural networks and how they came about and then dove into the idea of a perceptron and its basic form of functionality. We then embarked on a journey to explore four of the most common neural networks out there: MLP, CNN, RNN, and LSTM. We gained a better sense of how to select activation functions, measure loss, and implement our understandings using the Keras library.
Next, we took a less theoretical and much more hands-on approach as we tackled our first dataset that was sequential nature. We spent a considerable amount of time preprocessing our data, developing our model, getting our model development organized with MLflow, and reviewing its performance. Following these steps allowed us to create a custom and well-suited model for the problem at hand. Finally, we took a no-code approach by...