Summary

Computer vision is a big field within AI. By understanding this field, you can achieve results such as extracting information from an image or generating images that look just like they do in real life, for example. This chapter has covered image preprocessing for feature extraction using the OpenCV library, which allows easy training and prediction for machine learning models. Some basic machine learning models have also been covered, such as decision trees and boosting algorithms. These served as an introduction to machine learning and were mostly used to play around. Finally, neural networks were introduced and coded using Keras and TensorFlow as a backend. Normalization was explained and put into practice, along with dense layers, though convolutional layers are known to work better with images than dense layers do, and they will be explained later in the book.

Concepts for avoiding overfitting were also covered, and toward the end, we used the model to make predictions and put it into practice using real-world images.

In the next chapter, the fundamentals of natural language processing (NLP) will be introduced, along with the most widely used techniques for extracting information from a corpus in order to create basic models for language prediction.