Deep Learning and Artificial Intelligence
Professor Field Cady · Data Science · 2020
“Artificial Intelligence” (AI) has replaced Big Data as the buzzword of choice in describing data science. The key technology at the center of modern AI is “artificial neural networks,” a family of extremely powerful machine learning models that can do tasks that were previously unimaginable. This chapter is devoted to neural networks, including their strengths and weaknesses. In deep learning though, the whole point of a complicated neural network is that the early layers extract features from the raw data, and subsequent layers capture increasingly complex features that culminate in whatever it is readers are trying to predict. The chapter provides readers an overview of natural language processing technologies that blends deep learning with more traditional approaches. A key feature of knowledge bases is that – unlike neural networks that might “learn” information by baking it into a complicated mathematical hash – the content can be understood, fact-checked, and edited by humans.