Statistical Analysis

This course covers the ways in which analytics are being used in the healthcare industry. Topics include data collection opportunities created by the ACA and other laws, the use of analytics to prevent fraud, the use of predictive modeling based on medical records, the insurance industry's increasing use of data and the ethical issues raised by these practices. Prerequisites: none required (470.681 Probability and Statistics recommended)

In this course students will develop expertise in using the tools necessary to collect, analyze, and visualize large amounts of text. The course begins with a hands-on introduction to the programming concepts necessary to collect and process textual data. The course then proceeds to cover key statistical concepts in machine learning and statistics that are used to analyze text as data. Throughout the course, students will develop a research project that culminates in the display of results from a large-scale textual analysis. Prerequisite: 470.681 Probability and Statistics.

Machine learning and, more broadly, artificial intelligence, can now be used to perform complex tasks such as image recognition, fraud detection, traffic prediction and product recommendations. These successes are driven by developments in machine learning, such as the use of neural networks. This course introduces students to a variety of machine learning techniques using Python. Students will first learn the fundamentals of the Python programming language and will then implement machine learning algorithms and develop an understanding of how they work. Further, students will learn how to select and implement an appropriate algorithm depending on the type of dataset they have, and will be able to use a machine learning algorithm to generate predictions. Prerequisite: 470.681 Probability and Statistics

Data are everywhere, and many elected officials and government managers understand they need it. But how can they use it to solve problems and shape policy? What is the best way to make decisions based on a data analysis? How can they communicate those decisions, and the rationale behind them, to employees, citizens, and stakeholders? This course will provide students with an experiential learning opportunity based on real-world scenarios. Students will each take on a role (mayor, police commissioner, human capital director, budget director, public works director, public health director) and participate in a simulated public policy scenario. Working in small groups, students will apply a practical performance analytics process to develop solutions to address governmental challenges. Students will begin by studying foundational concepts and techniques of data collection, analytics, and decision support. They will also learn how to navigate multiple interests, asymmetrical information, and competing political agendas as they make difficult decisions about resource allocation and public policy. Along the way, they will learn how to turn insights into action by effectively communicating the results of analysis to busy executives and decision makers at all levels of the organization. Prerequisites: none required (470.681 Probability and Statistics recommended)

This course provides insights into how to utilize shared cloud computing resources through a service provider. These resources can be storage space, software as a service, or compute servers. This is a hands-on course in which students will access a variety of cloud services and work with different cloud providers such as Apple, Microsoft, Google, and Amazon. Students will set up virtual servers, work with cloud file storage, learn about a variety of cloud collaboration options, and much more. This practical course will help students make the transition to working in the cloud from any device, anywhere, anytime. All areas of the public sector, such as education, healthcare and law enforcement, increasingly use cloud computing both to deliver information to clients and share information within and across agencies.

This class applies data analytic skills to the urban context, analyzing urban problems and datasets. Students will develop the statistical skills to complete data-driven analytical projects using data from city agencies, federal census data, and other sources, including NGOs that work with cities. We will examine a variety of data sets and research projects both historical and contemporary that examine urban problems from a quantitative perspective. Over the course of the term, each student will work on a real-world urban data problem, developing the project from start to finish, including identifying the issue, developing the research project, gathering data, analyzing the data, and producing a finished research paper. Prerequisite: 470.681 Probability and Statistics

Learning the basics of Python empowers analysts to retrieve and leverage data in new ways. After covering the fundamentals of syntax, students learn how to read, create and edit data files using Python. Building on that knowledge, students interact with online resources through bulk data APIs and web scraping. Finally, students will use the data they collect to develop an original analysis. Prerequisite: 470.681 Probability and Statistics

This course addresses the legal, policy and cultural issues that challenge the government and its citizens in the increasingly complex technical environment of privacy. We will examine the challenges in balancing the need for information and data against the evolving landscape of individual privacy rights. The course will examine privacy at all levels: by analyzing the shifting views of individual privacy by citizens as well as the technological challenges in both protecting and analyzing personal information for government use. Using case studies and hypotheticals, we will discuss the issue of transparency in the government use and retention of data. The cases will range from Facebook to healthcare.gov to sunshine laws to national security uses of information. We will trace the development of legal and policy measures relevant to privacy concerns and envision future solutions needed in an era of great technological innovation including the use of big data. Prerequisite: none

This course explores technological and data-driven solutions for policy challenges. This includes developments within government, such as the new types of leadership provided by Chief Innovation or Chief Data Officers, the trend toward digitalization of services, and the movement toward open data. It also covers innovation by citizens through the civictechnology movement. Civic tech initiatives have been used to extend and improve services, increase efficiency, design applications for citizen engagement, and improve communication across a variety of policy domains. The course also covers the concept of smart cities and how it can be understood as both new applications of technology (such as sensors and smart infrastructure), and the strategic use of data. For the course project, students will evaluate a policy initiative using city open data, policy research, an analysis of political culture within which the initiative would be implemented, and the technology that could be used for the initiative. Some familiarity with R programming language and theRStudio environment is helpful. Prerequisites: one of the following: 470.681 Probability and Statistics or 470.768 Programming and Data Management

Many government agencies engage in data mining to detect unforeseen patterns and advanced analytics (such as classification techniques) to predict future outcomes. In this course, students will utilize IBM SPSS Modeler to investigate patterns and derive predictions in policy areas such as fraud, healthcare, fundraising, human resource and others. In addition, students will build segmentation models using clustering techniques in an applied manner. Integration with other statistical tools and visualization options will also be discussed. Prerequisite: 470.681 Probability and Statistics; Recommended: 470.709 Quantitative Methods

Analytics inform the decision-making process, strategizing, and forecasting of modern American campaigns. This course focuses on the role that analytics play in campaigns and elections in America. Campaign strategists, policy analysts, and social scientists leverage data from voter rolls, consumption and public opinion polls to make better choices. This course surveys the theoretical and empirical literature in American electoral politics to examine how campaigns and political organizations are using field experiments, microtargeting, and public opinion polling to tackle the challenges of getting out the vote and increasing registration and voting rates. Other topics covered include voting behavior, public opinion, partisanship, and campaign finance. Students will gain a rich understanding of how analytics has become a key component of the electoral process. Students will also gain experience analyzing data through simulations and data analysis exercises. Prerequisites: none required (470.681 Probability and Statistics recommended)

This course provides students with a strong foundation in database architecture and database management systems. Students will evaluate the principles and methodologies of database design and techniques for database application development. Students will also examine the current trends in modern database technologies such as Relational Database Management Systems (RDBMS), NoSQL Databases Cloud Databases, and Graph Databases. Prerequisite: none

This course is a comprehensive examination of all aspects of designing questionnaires, conducting survey research, and analyzing survey data. The class will cover question construction, measurement, sampling, weighting, response quality, scale and index construction, IRBs, ethics, integrity and quality control, modes of data collection (including telephone, mail, face to face and focus groups), post collection processing and quantitative analysis of data (including chi-square and ANOVA), as well as report writing fundamentals. The class culminates by fielding a survey of student created questions and writing an executive summary of the survey with a paper discussing the research findings. Prerequisite: 470.681 Probability and Statistics

Data science is a methodology for extracting insights from data. This course is an introduction to the concepts and tools that are used in data science with an emphasis on their application to public policy questions. The course covers some advanced data mining and machine learning processes including classification and decision trees, random forests, cluster analysis, and outlier detection, while also providing you with training in the basics of data management and data exploration. All of the work in the course will be conducted to prepare you to proficiently conduct predictive analytics in a real-world setting. Some familiarity with R programming language and the RStudio environment is necessary. Prerequisite: 470.681 Probability and Statistics

This course will address the practical applications of artificial intelligence particularly in the realms of policy and governance. AI and data science are transformational technologies that hold the promise of improving lives and society at large. While excitement about AI and its applications is growing, its adoption is anything but straightforward. The successful application of AI to lower risk, better understand customers and automate decision making requires a deep knowledge of the right use cases where AI can lead to breakthrough innovations.

This course will provide students with the opportunity to investigate multiple AI use cases and evaluate their merit. In addition, students will select a specific use case, develop reference architecture and determine an appropriate implementation strategy. The course will culminate in the development and delivery of a lab-to-market strategy for their selected use case. No prerequisite.

This course will introduce computational modeling and demonstrate how it is used in the policy and national security realms. Specifically, the course will focus on agent-based modeling, which is a commonly-used approach to build computer models to better understand proposed policies and political behavior. Agent-based models consist of a number of diverse "agents,'’ which can be individuals, groups, firms, states, etc. These agents behave according to behavioral rules determined by the researcher. The interactions with each other and their environment at the micro-level can produce emergent patterns at the macro-level. These models have been used to understand a diverse range of policy issues including voting behavior, international conflict, segregation, health policy, economic markets, ethnic conflict, and a variety of other policy issues. The course will consist of two parts: First, we will examine the theoretical perspective of computational modeling. Second, you will be introduced to a software platform that is commonly used to develop computational, and, in particular agent-based modeling. No prerequisite

Washington, D.C. is the laboratory for anyone studying American government and politics or analyzing the policy making process here. DC Lab: Politics, Policy, and Analytics will give any graduate student in one of the programs of the JHU Center for Advanced Governmental Studies the opportunity to bring theory and practice together through an intensive week of lectures, seminars, and site visits in the nation’s capital. Sessions will include guest speakers from JHU faculty, think tank scholars, and agency officials. The goal is to experience Hopkins in Washington and assess what is observed to better inform each student’s studies of the political process. Special Note: This course will require one week of residency in Washington, D.C. for the week of May 12-17, 2019.All travel and accommodations (food and lodging) in DC will be covered by the student. In addition to the course tuition, a $300 course fee will be charged to help cover costs for course incidentals.

Spatial Statistics is a rapidly developing tool in the discipline of ecology that analyzes both 2-D and 3-D data that contain a spatial component. Many ecologists use continuous data (e.g., vegetation density and height, net aboveground primary production, percent of biomass killed by disturbance, etc…) that violates the assumption of spatial independence; therefore, necessitating the need to analyze the data using spatial statistics. Thus, spatial statistics provides concepts, tools, and approaches that will enhance the analyses of population data, sample data, partitioning of regions (patch and boundary), spatial interpolation, and data that are spatially autocorrelated. The goal of this course is to give students a firm grasp of the concepts of spatial statistics in ecology and of how they can be applied to analyze continuous data for environmental policy, management, and assessment. Uses of case studies, data analysis in the R spatial statistics package, and discussions help to examine and apply the concepts.