Data Mining and Analytics

Data Mining and Analytics introduces students to the practical fundamentals and emerging paradigms of data mining and machine learning with enough theory to aid intuition building. The course is project oriented, with a project beginning in class every Thursday and to be completed outside of class by the following week, or two for longer assignments. The in-class portion of the project is meant to be collaborative and a time for the instructor and GSIs to work closely with project groups to understand the objectives, help  work through software logistics, and connect project work to lecture. Tuesday lectures introduce theories, concepts, contexts and algorithms. Students should expect to leave the class with hands-on, contemporary data mining skills they can confidently apply in research and industry. There will be a written midterm test and a final group project report and presentation. Experience with Python is required.

Learning Objectives

• Foster critical thinking about real world actionability from analytics.

• Develop intuition in various machine learning classification algorithms (e.g., decision trees, feed-forward neural networks, recurrent neural networks, support vector machines) and clustering techniques (e.g., k-means, spectral, skip-gram)

• Conduct manual feature engineering (from domain knowledge) vs. machine induced featurization (representation learning)

• Provide an overview of issues in research and practice that will affect the practice of data science in a variety of domains.

This course is currently offered as Info 154.