A Quick Primer on Deep Learning, Reinforcement Learning, and Artificial Intelligence

DEAN OF BIG DATA TIP:

While it is unlikely that you will ever be asked to build your own neural network or RL algorithm, it is important to understand how these advanced analytics work (at a high level) and what can be done with them from a value creation perspective. These are the tools of a modern-day value creation alchemist.

DL is a set of algorithms that analyze massive datasets using a multi-layered neural network structure, where each layer is comprised of numerous nodes, to train and learn to recognize and codify patterns, trends, and relationships buried in the data… without human intervention. Common applications of DL include image recognition, natural language processing, disease detection, and facial recognition (see Figure 6.4).

Figure 6.4: How Deep Learning Works

There are two key capabilities that underpin the continuous learning nature of...