This course introduces students to statistical and machine learning tools for modeling unstructured data; including emails, documents, text messages, and social media data. Topics to be covered include text mining, clustering, mixture models, deep learning, and topic models. The course integrates numerous applications to demonstrate practical approaches to analyzing large unstructured collections of data. Application areas include Marketing (Yelp and Trip Advisor reviews), Human Resources (health care plan analysis), Social Media (Twitter, YouTube, and Instagram). The course delivery will be a mix of lectures, readings/podcasts with discussion, and hands-on data analysis. Prereq: DS 805.

The course is designed to introduce analytical techniques needed in the estimation and analysis of temporal (time series) data in various business disciplines. The course focuses on theoretical and application aspects of stationary/non-stationary univariate as well as multivariate time series models. Emphasis will be given to topics such as time series regression, random walks, ARIMA/SARIMA processes, ARCH/GARCH for modeling conditional volatility, Vector ARMA models, and transfer functions. Modern software implementation is fully integrated into the course. Some examples of the business application areas include demand forecasting, financial asset return modeling, stochastic volatility modeling of financial indexes and securities, mortgage default risk assessment, online webpage click-rate modeling, market share modeling. Prereq: DS 803.

Special Topics; may be repeated. Pre- and co-requisite courses vary. Please consult time and room schedule for the specific 898 topics section you are interested in for details.

Fundamental analysis and design principles. Number systems, codes, Boolean algebra, and combinational and sequential digital circuits. Lab: student-built systems using modern integrated circuit technology and an introductory design session on a CAD workstation. Lab.