Machine Learning in Practice

Catalog description

Machine Learning is concerned with computer programs that enable the behavior of a computer to be learned from examples or experience rather than dictated through rules written by hand. It has practical value in many application areas. This class is meant to teach the practical side of machine learning for applications, such as mining newsgroup data, building adaptive user interfaces or building natural language processing applications. There will be a significant project focus, and when you have completed the course, you should be fully prepared to attack new problems using machine learning. While it will be essential to learn conceptually how machine learning algorithms work and interact with data, the emphasis will be on effective methodology for using machine learning to solve practical problems. This is about knowing how to conceptualize a problem, knowing how to represent your data, being able to interpret your results properly, doing an effective error analysis, and using the results of the error analysis to make strategic decisions about how to adjust the way you have set up your data and selected and tuned your algorithms. We will cover a wide range of learning algorithms that can be applied to a variety of problems. In particular, we will cover topics such as decision trees, rule based classification, support vector machines, Bayesian networks, clustering and neural networks. In addition to readings from the course textbook, we will have additional readings from research articles that will be announced ahead of time and distributed on Canvas. In the last third of the course, there will be an introductory and non-technical coverage of the main concepts in natural language processing (NLP) and advances in machine learning, covering topics such as neural networks and deep learning, word embeddings, large language models and some of their applications.