Machine Learning and Deep Learning within the Context of Artificial Intelligence

This chapter provides an overview of the connections between Artificial
Intelligence (AI), Machine Learning (ML), and Deep Learning (DL), offering a broader
view of how ML and DL fit within the overall AI landscape. In this chapter, we explore
the historical roots of AI in philosophy before discussing its eventual development into
both mathematical constructs and programming. There remains considerable confusion
regarding the definitions of AI, ML, and DL. While many refer to any recent ML or DL
model as "AI," this is not entirely accurate. The primary branches of thought are
symbolic representation (using logic-based rule sets), ML (data-driven learning), and
DL (complex networked layers used for data processing). We briefly discuss
algorithms and architecture and delve into the underlying mathematics. Using several
domains such as CV, NLP, Healthcare, and Recommender Systems as examples, we
illustrate how these ideas have transformed the way machines operate. However, we
also emphasize the importance of fairness, interpretability, and ethics, as well as the
need to understand how these systems operate. Finally, we conclude with a discussion
of the future direction of AI, including Neurosymbolic AI, Explainable Models, and
Edge Computing. We advocate that AI should continue to be grounded in human
values and sensibility despite its growing complexity. The aim is to clarify the
connections between AI, ML and DL. While complete clarity is difficult to achieve
given the complexity of this area, we highlight the most important concepts to establish
the key distinctions and similarities between AI, ML and DL.