Part 3: Design and Development of ML Systems
Although machine learning and its cousin artificial intelligence are well known, they refer to a wide range of algorithms and models. First, there are the classical machine learning models that are based on statistical learning and usually require the data to be prepared in a tabular form. They identify patterns in the data and can replicate these patterns. However, there are also modern models that are based on deep learning, which are capable of capturing more fine-grained patterns in data that is less structured. The crown examples of these models are the transformer models (GPT) and autoencoders (diffusers). In this part of the book, we take a closer look at these models. We focus on how these models can be trained and integrated into machine learning pipelines. We also look into how software engineering practices should be applied to these models.
This part has the following chapters:
- Chapter 9, Types of Machine Learning...
