Master of Applied Data Science

When you earn your MADS degree, you’ll be able to:

• Use data to improve outcomes and achieve ambitious goals.
• Master core data science concepts like machine learning and natural language processing.
• Understand key topics such as privacy, data ethics, and persuasive communication.
• Build predictive models and visualize data in dashboards.
• Formulate better problem statements for data informed solutions.

Curriculum

Students enrolled in the University of Michigan School of Information’s Master of Applied Data Science (MADS) program will take courses in all essential subjects of applied data science, with an emphasis on an end-to-end approach. The MADS program intersects computation with theory and application, ensuring that students put their data science learnings into practice.

Courses cover:

• Computational methods for big data
• Exploring and communicating data
• Visualizing data using various methods
• Analytic techniques (machine learning, network analysis, natural language processing, experiments and causal inference)
• Data science applications in context (search and recommender systems, social media analytics, learning analytics)
• 3 portfolio-building major projects

The following course clusters and titles highlights a breadth and depth of engaging data science subjects. Courses cover everything from problem formulation to putting results into action.

Python is the primary programming language used throughout this curriculum. Students will apply data science skills and knowledge in 3 capstone projects throughout the program.

Unless otherwise noted, each course is 1 credit unit (roughly 4 weeks) in length. A total of 34 credit units is required to graduate. Please note that course titles are subject to change as the curriculum is expanded and refined.

Formulating Problems

• Introduction to Applied Data Science
• Contextual Inquiry
• Data Science Ethics

Collecting and Processing Data

• SQL & Databases
• SQL Architectures & Technologies
• Big Data: Efficient Data Processing
• Big Data: Scalable Data Processing
• Data Manipulation
• Experiment Design and Analysis

Analyzing and Modeling Data

• Math Methods for Data Science
• Visual Exploration of Data
• Data Mining I
• Data Mining II
• Supervised Learning
• Unsupervised Learning
• Deep Learning
• Machine Learning Pipelines
• Causal Inference
• Natural Language Processing
• Network Analysis

Presenting and Integrating Results into Action

• Information Visualization I
• Presenting Uncertainty
• Communicating Data Science Results
• Information Visualization II

Real world applications of data science

• Search and Recommender Systems
• Social Media Analytics
• Learning Analytics
• More to come

Culminating Learning Experiences

• Capstone I: synthesis of computational techniques to collect and process big data
• Capstone II: synthesis of analytics and machine learning techniques to analyze data and present results
• Capstone III: capstone that applies end-to-end data science techniques to real world scenarios

MADS students have the opportunity to start with these data science courses:

Introduction to Applied Data Science

This course explores expertise, perspectives, and ethical commitments data scientists apply to projects during four phases of data science: problem formulation, data acquisition, modeling and analysis, and presentation of results. Through this process, students will define a vision for how they want their data science careers to develop.

Data Manipulation

Data Manipulation presents manipulation and cleaning techniques using the popular Python Pandas data science library. By the end of this course, students will have the skills needed to take tabular data, clean it, manipulate it, and run basic inferential statistical analyses.

Math Methods for Data Science

Math Methods will review and establish the foundational math concepts needed for a data scientist’s toolkit. Students will learn and apply concepts from linear algebra (such as matrices and vectors), basic optimization techniques (such as gradient descent), and statistics (such as Bayes’ rule).

Information Visualization I

Information Visualization I will focus is on the role of visualization in understanding one-dimensional and multidimensional data. It covers how perception, cognition, and good design can enhance visualizations. This course also introduces APIs for visualization construction.

Experiment Design and Analysis

Experiment Design and Analysis presents techniques for laboratory and field experiments. Students will discuss the logic of experimentation and the ways in which experimentation is used to investigate social and technological phenomena. Students will also learn ways to design experiments and analyze experimental data.

Visual Exploration of Data

Visual Exploration of Data enables students to identify aggregate patterns within data using the matplotlib library, and learn the challenges associated with exploring and representing data. Students will also improve their understanding of the applications of various statistical methods.

Data Mining I

Data Mining I introduces the basic concepts of data mining. This course covers how to represent real world information as basic data types (itemsets, matrices, and sequences) that facilitate downstream analytics tasks. Students will learn how to characterize each type of data through pattern extraction and similarity measures.