Neural network architectures as graphs

In Chapter 1, we touched on deep learning models, particularly in the context of generative artificial intelligence. Deep learning models surface in many areas of analytics, including natural language processing, image classification, and time series forecasting. Let’s explore these applications in more detail.

Natural language processing is ubiquitous in the era of big data. Surveys often employ free text collection methods. Customers provide text reviews of products. Social scientists jot down notes when they are observing populations qualitatively (dubbed ethnographic research). Bloggers post regular content to disseminate ideas. Legal, medical, and educational notes can include biased content that needs to be addressed to protect people from institutional bias that limits future opportunities and wellness.

All this data needs to be parsed before it is fed into a classification model or exploratory data tools. Most of the tools...