Data Science Master's

Teaches mathematical and statistical fundamentals for data science and focuses on the modern computational statistical modeling methods. Covers basic computational statistical inference, the properties and behavior of linear models, conditional probability, Bayesian models, inference by resampling and time series methods.

Covers comprehensive and practical approaches to research, including specific methods of analysis for students to develop advanced research skills in the general areas of descriptive statistics, exploratory data analysis and confirmatory data analysis. Includes methods to better communicate results of the research. Successful students will have the skills to be useful participants in the data lifecycle from collection to management, analysis and visualization of results.

Hands-on​ ​introduction​ ​to​ ​deep​ ​learning emphasizing application using GPU-accelerated hardware to train multilayer machine learning models directly on raw input signals. Discuss​es ​the foundations​ ​of​ ​feedforward​ ​networks,​ ​convolutional​ ​neural networks,​ ​and​ ​recurrent​ ​networks,​ ​as​ ​well​ ​as​ ​their usage​ ​within​ ​popular​ ​reinforcement​ ​learning​ ​frameworks.​ ​​Using​ ​real​ ​datasets​ ​and​ popular deep learning tools (e.g. Tensorflow, Keras) students create systems to make inferences from rich and varied raw data including speech, video and other sensor signals.

Covers major topics used in developing predictive modeling and applied statistical forecasting models that are of major interest to business, government and academia. These include exploring the calibration of models, the estimation of seasonal indices and the selection of variables to generate operational business forecasts. Topics assist business professionals in utilizing historical patterns to build a more constructive view of their future. Overview of how these topics can be used with data capture, integration and information deployment capabilities to ensure more productive decisions and more accurate planning. Modern forecasting techniques are covered for the evaluation of sophisticated business models used to make intelligent decisions in marketing, finance, personnel management, production scheduling, process control, facilities management and strategic planning.

Insightful displays of complex, large and possibly unstructured quantitative and qualitative data. Data visualization for analytics goes beyond traditional static graphs and charts by seamlessly connecting data analysis, data-based optimization and data presentation to create visualizations. Topics include visualization design principles, data refinement and preparation, tandem modeling and optimization with visualizations, use of state-of-the-art software tools for visualization and creation of dynamic interactive visualizations as decision support aids. A semester project in data visualization for analytics relevant to a functional area of business is an important component of the course.