Curriculum

This course provides general exposure to basic statistical concepts that are necessary for students to understand the content presented in more advanced courses in the program. The course covers theoretical distributions and the way these distributions are used to assign probabilities to events in some depth. The course also introduces students to descriptive statistical methods to explore and summarize data, methodologies for sampling units for measurement or analysis, drawing inferences on the basis of knowledge gained from samples to populations, assessing relationships between variables, and making predictions based upon relationships between variables.

If you are required to take this course it will be held the 5 weeks leading up to the start of your first quarter.

This course is an introduction to the essential concepts and techniques for the statistical computing language R. Topics covered include the R and RStudio environment, arithmetic, basic data structure, importing and exporting data, visualization, and basic statistics.

If you are required to take this course it will be held concurrently with the second five weeks of your first quarter in the program.

This course in Python starts with introduction to the Python programming language basic syntax and environment. It methodically builds up the learner's experience from the level of simple python statements and expressions to writing succinct, efficient, and fast Python expressions and package the code in methods and classes. In general, the course is geared toward developing a data scientist's toolbox such as data importing, cleaning and preparation, and covers a number of machine learning algorithms. The course expands beyond these skills as it stresses upon the importance of some of Python's most unique and powerful features and serves as an introduction to object oriented programming and Python Classes.

If you are required to take this course it will be held concurrently with the first five weeks of your first quarter in the program.

The advanced linear algebra course is focused on the theoretical concepts and real-life applications of linear algebra for machine learning. Upon completion of this course, students are going to have a strong foundation of linear algebra and linear analysis topics essential for the development of core machine learning and data mining concepts.

For example, airline companies employ time series to predict traffic volume and schedule flights; financial agencies measure market risk via stock price series; marketing analysts study the impact of a newly proposed advertisement by the sales series. A comprehensive knowledge of time series analysis is essential to the modern data scientist/analyst. This course covers important issues in applied time series analysis: a solid knowledge of time series models and their theoretical properties; how to analyze time series data by using mainstream statistical software; practical experience in real data analysis and presentation of their findings in a logical and clear way to various audiences.

Prerequisite: 

• MSCA 31007: Statistical Analysis
• Completion of 400 units in the MScA program

Throughout the course, students will learn concepts and fundamentals of statistical inference and regression analysis by studying theory, developing intuition, and working through several practical examples. Students will become proficient in interpreting standard regression output and conducting model selection and validation. Students will also learn the statistical programming language used to construct examples and homework exercises. Examples will be constructed using SAS or R. Students will have many opportunities to apply the new concepts to real data and develop their own statistical routines. The course also addresses the importance of quality control and reproducibility when conducting research and developing work product.

Prerequisites:

• MSCA 31000: Introduction to Statistical Concepts
• MSCA 37016: Advanced Linear Algebra for Machine Learning

This course will introduce students to the common algorithms: association and sequence rules discovery, memory-based reasoning, clustering, classification and regression decision trees, logistic models, and neural network models. In addition, students will learn how to compare analytical results and give recommendations during the data mining process. In addition to assignments and tests, students are required to propose and complete a data mining research project of their own design.

Prerequisite:

• MSCA 31007: Statistical Analysis

It will emphasize practice over mathematical theory, and students will spend a considerable amount of class time gaining experience with each algorithm using existing packages in R, Python, and Linux libraries. The course will cover the following topics: regression and logistic regression, regularized regression including the lasso and elastic net techniques, support vector machines, neural networks, decision trees, boosted decision trees and random forests, online learning, k-means and special clustering, and survival analysis.

Prerequisites:

• MSCA 31008: Data Mining Principles 
• MSCA 31010: Linear & Non-Linear Models 
• MSCA 37014: Python for Analytics

Course emphasizes applications of these models to various fields and covers main steps of building analytics from visualizing data and building intuition about their structure and patterns to selecting appropriate statistical method to interpretation of the results and building analytical model. Topics are illustrated by data analysis projects using R. Familiarity with R at some basic level is not a requirement but recommendation. Students can pick up the programming language by following the descriptions of the examples.

Prerequisite: 

• MSCA 31007: Statistical Analysis

Modern data engineering platforms reduce manual data preparation by automating processes, which in turn, enable companies to focus on deriving efficiencies in data processing to develop impactful business insights. This course provides students with a thorough understanding of the fundamentals of data engineering platforms, for both operational and analytical use cases, while gaining hands-on expertise in building these platforms in a way to develop analytical solutions effectively. Students will have the opportunity to construct both relational and analytical databases on the cloud or on premise from real-life datasets while using programmatic or configuration driven data pipelines.

By the end of the course, students will be able to design and implement an end-to-end data engineering platform capable of supporting sustainable analytics solutions.

Students will learn the big data infrastructure including Linux, massive parallelization, and distributed computing, and how to apply both Hadoop and Spark map-reduce concepts for clustering, similarity search, web analytics and classification. During the course, we will cover the applications of NoSQL systems, such as JSON stores, object storage and Elasticsearch. The cloud computing section of the course will focus on virtualization and container orchestration, including virtual machines, dockers and Kubernetes. During the course students will gain hands-on expertise leveraging Hive, Pig, Python and PySpark for big data applications in client-server environment.

MScA students in the 12-course curriculum who wish to take Big Data Platforms as a core course, instead of MSCA 31012 Data Engineering Platforms, should have working knowledge of the following prior to taking Big Data Platforms:

• Database Design
  • Relational databases (normalization, de-normalization)
• Familiarity with relational databases and SQL, including the following statements:
  • CREATE TABLE
  • SELECT
  • FROM
  • WHERE
  • GROUP BY
  • ORDER BY
• Python 101, including the following libraries:
  • Pandas, Matplotlib
  • Familiarity with scikit-learn and XGBoost is beneficial, but not required.
• Foundational knowledge of Linux:
  • Students without Linux knowledge will need to take a free online bootcamp provided by MScA program

Suggested skills and knowledge:

• Distributed computing
  • Multithreading
  • CAP Theorem
• No-SQL database design
  • Document databases
  • Graph databases {applied to graph compute in Spark)
  • Key value stores
• Cloud Platforms
  • IaaS/PaaS/SaaS
  • List of services and access methods
  • Resource Provisioning / Scripting
  • Programmatic Authentication
  • Cloud APIs

The goals of the course are

1. to identify points in an organization that can benefit from analytics;
2. to structure analytic problems from a strategic perspective, thereby identifying business impact;
3. to develop the ability to communicate the power of analytics to others, especially senior leaders; and
4. to work in a team to accomplish these and related goals successfully.

At the end of the course, students should have the ability to describe business problems that lend themselves to a data analytics approach, position these problems from the perspective of a coherent business strategy, and represent the power of analytics to a business audience. Students should also understand how to harness the powerful dynamics of a team to achieve excellence in the world of data analytics.

The successful data science leader / consultant brings an uncommon combination of deep business acumen, data literacy, leading edge methodology experience, inspirational team leadership, client communication management and organizational change skills.

Successful consultants rely on a variety of consulting tools to diagnose organizational problems, identify solutions and deliver those solutions.

The Data Science for Consulting course will enable students:

• Understanding the structure of consulting organizations and engagements.
• Developing data science solutions to enterprise problems through employing traditional consulting frameworks and best practice tools.
• Practicing successful project delivery through effective data discovery, communication, influential team leadership and client relationship management.

The course puts special emphasis on covering main steps of building analytics from visualizing data and building intuition about their structure and patterns to selecting appropriate statistical method to interpretation of the results and building analytical models. Topics are illustrated by data analysis projects using R. 

Prerequisites: 

• MSCA 31007: Statistical Analysis
• Basic familiarity with R is a requirement.

We will examine data requirements and sources of data; utilize statistical techniques and visualization methods to evaluate data completeness and quality; assess and compare model performance; learn how to effectively communicate analytical insights to non-technical audience. Students will learn through a combination of in-class discussions, case studies, and team projects.

Team exercises will teach students effectively communicate between business process owners and analytical experts to overcome typical barriers in Business Analytics, such as data availability, resource constraints, and resistance to change.

We will cover such topics as security concerns unique to the field, research design strategies, and the integration of epidemiologic and quality improvement methodologies to operationalize data for continuous improvement. Students will be introduced to the application of predictive analytics to healthcare. Students will understand factors impacting the delivery of quality and safe patient care and the application of data-driven methods to improve care at the healthcare system level, design approaches to answering a research question at the population level, become familiar with the application of data analytics to impacting care at the provider level through Clinical Decision Systems, and understand the process of a Clinical Trial.

Prerequisite:

• MSCA 31007: Statistical Analysis

First class has been developed to optimize real, non- simulated systems or to find the optimal solution of a mathematical model. The methods that belong to this class include liner programming, quadratic programming and mixed-integer programming. Second class of methods has been developed to optimize a simulation model. The difference with the classical mathematical programming methods is that the objective function (which is the function to be minimized or maximized) is not known explicitly and is defined by the simulation model (computer code).

The course will demonstrate multiple approaches to build simulation models, such as discrete event simulations and agent-based simulations. Then, it will show how stochastic optimization and heuristic approaches can be used to analyze the simulated system and design a sequence of computational experiments that allow to develop a basic understanding of a particular simulation model or system through exploration of the parameter space, to find robust plausible behaviors and conditions and robust near-optimal solutions that are not prone to being unstable under small perturbations.

Prerequisite:

• MSCA 31007: Statistical Analysis

The focus of this course is an introduction to Bayesian approach. Many methods learned by students in Statistical Analysis, Linear and Nonlinear Models, Data Mining and Machine Learning will be reviewed from the point of view of probabilistic inference. We will look at hierarchical, mixture, robust, and non-parametric Bayesian models and learn how to use them in practical applications. Content will include using probabilistic models to make business decisions under uncertainty, analyze causation in the data, use probabilistic inference to assess risk of black swan events, account for uncertainty in project management and other applications.

Students will learn necessary facts of probability theory, Bayesian reasoning, Markov chain Monte Carlo using JAGS, STAN and PyMC. The course contains large number of interactive demonstrations, workshops with examples through which the lecturer shares his own hands-on experience with the students.

Prerequisite:

• MSCA 31010: Linear and Nonlinear Models

The goal is to provide marketers with the foundation needed to apply data analytics to real-world challenges they confront daily in their professional lives. Students will learn to identify the web analytic tool right for their specific needs; understand valid and reliable ways to collect, analyze, and visualize data from the web; and utilize data in decision making for their agencies, organizations or clients.

By completing this course, students will gain an understanding of the motivations behind data collection and analysis methods used by marketing professionals; learn to evaluate and choose appropriate web analytics tools and techniques; understand frameworks and approaches to measuring consumers’ digital actions; earn familiarity with the unique measurement opportunities and challenges presented by New Media; gain hands-on, working knowledge of a step-by-step approach to planning, collecting, analyzing, and reporting data; utilize tools to collect data using today’s most important online techniques: performing bulk downloads, tapping APIs, and scraping webpages; and understand approaches to visualizing data effectively.

But the biggest challenge of all is learning to think differently in order to ask new types of questions that could not be answered by analyses of less complex data streams with less complex technological infrastructure. In recent years significant progress has been achieved in creating technological ecosystems for big data analysis. Innovative technologies such as open source projects MapReduce, Hadoop, Spark, Storm, Kafka, TensorFlow, H2O, etc. allowed us to look at depths of data unseen before. We have now growing number of sources and educational courses introducing these new tools.

It appeared little more difficult to develop new data analysis methods appropriate for the new data ecosystems. There are some new interesting ideas, there is significant amount of empirical studies. But the methods of late 19^th and first half of 20^th centuries, albeit transformed to run with technology of 21^st century, are still dominating our research. Traditional and new concepts and methods of big data analysis are typically, with few exceptions, covered in books and taught in courses requiring a laptop environment, often testing limits of computing power of personal hardware, but not giving enough flavor of the 21^st century combination of high-end technology and discussion of methods in depth.

The goal of this course is to fill this gap and teach students to think about real problems by analyzing big data in new data analysis ecosystems. The course is project-based: we will take up several real life projects and discuss different approaches to digging for insights, possible pitfalls and applications. In our work on projects we will use Python and modern cloud computing environments: Spark, TensorFlow.

Prerequisites:

• MSCA 31009: Machine Learning & Predictive Analytics 
• MSCA 37011: Deep Learning & Image Recognition (Recommended)

Students in this course will learn foundations of natural language processing, including: concept extraction; text summarization and topic modeling; part of speech tagging; named entity recognition; semantic roles and sentiment analysis. For advanced NLP applications, we will focus on feature extraction from unstructured text, including word and paragraph embedding and representing words and paragraphs as vectors. For cognitive analytics section of the course, students will practice designing question answering systems with intent classification, semantic knowledge extraction and reasoning under uncertainty. Students will gain hands-on expertise applying Python for text analysis tasks, as well as practice with multiple IBM Watson services, including: Watson Discovery, Watson Conversation, Watson Natural Language Classification and Watson Natural Language Understanding.

Prerequisite:

• It is strongly recommended that MSCA 31008 Data Mining Principles be taken before this course, but it is not a requirement to enroll.

Building such systems requires proficiency in programming, understanding of computer systems, as well as knowledge of related analytical methodologies, which are the skills that this course aims to teach to students. The course focuses on python and is tailored for students with basic programming knowledge in Python. The course is partially project based. During the first three sessions, we will review basic python concepts and then learn more advanced python and the ways to use Python to handle large data flows. The later sessions are project based and will focus on developing end-to-end analytical solutions in the following areas: Finance and trading, blockchains and crypto-currencies, image recognition, and video surveillance systems.

Prerequisites: 

• MSCA 31007: Statistical Analysis 
• MSCA 37014: Python for Analytics (Recommended)

Reinforcement learning refers to a system or agent interacting with an environment and learning how to behave optimally in such environment. An environment typically includes time, actions, states, uncertainty and rewards. Reinforcement learning combines neuro networks and dynamic programming to find an optimal behavior or policy of the system or agent in complex environment setting. Neuro networks approximations are used to circumvent the well-known 'curse of dimensionality' which have been a barrier to solving many practical applications. Dynamic programming is the key learning mechanism that the system or the agent uses to interact with the environment and improve its performance. Students will master key learning techniques and will become proficient in applying these techniques to complex stochastic decision processes and intelligent control.

Prerequisites:

• MSCA 31007: Statistical Analysis 
• MSCA 32013: Optimization and Simulation Methods for Analytics (Recommended)

This course crosses the chasm that separates machine learning projects/experiments and enterprise production deployment. It covers 3 pillars in MLOps: software engineering such as software architecture, Continuous Integration/Continuous Delivery and data versioning; model engineering such as AutoML and A/B experimentation; and deployment engineering such as docker containers and model monitoring. The course focuses on best practices in the industry that are critical to enterprise production deployment of machine learning projects.

Having completed this course, a student understands the machine learning lifecycle and what it takes to go from ideation to operationalization in an enterprise environment. Furthermore, students get exposure to state-of-the-art MLOps platforms such as Allegro AI, xpresso, Dataiku, LityxIQ, DataRobot, AWS Sagemaker, and technologies such as gitHub, Jenkins, slack, docker, and kubernetes.

Prerequisite: 

• MSCA 31009: Machine Learning & Predictive Analytics

This course in Deep Learning and Image Recognition will provide a practical, hands-on set of lectures on Deep Learning and Image Processing tools and techniques.

It will emphasize practice over advanced mathematical theory, and students will spend a considerable amount of class time gaining experience on Neural Networks and their applications in Python and other open source libraries.

The course would cover algorithms for undertaking competitive analysis in the digital landscape, market segmentation, mining databases for effective digital marketing, design of new digital and traditional products, forecasting sales and product diffusion, real time product positioning, intra omni-channel optimization and inter omni-channel resource allocation, and pricing across both omni-channel marketing effectiveness and ROI. The course will use a combination of lecture, in-class discussions, group assignments, and a final group project. The course lays special emphasis on algorithms. Hence it draws heavily from the fields of optimization, machine-learning based recommendation systems, association rules, consumer choice models, Bayesian estimation, experimentation and analysis of covariance, advanced visualization techniques for mapping brand perceptions, and analysis of social media data using advanced NLP techniques.

Prerequisites:

• MSCA 31007: Statistical Analysis

"Big Data" continues to grow exponentially in our large-scale transactional world where 100,000s of SKUs and millions of customers are interacting with 1:1 offers that include differential pricing, shipping timing/costs and even made to order "custom" product configurations. These consumer behaviors are quickly advancing the availability of new data and techniques within the discipline of Data Science. This elective course will give students the opportunity to apply their skills in data visualization, data mining tools, predictive modeling, and advanced optimization techniques to address Supply Chain challenges. Students will also learn how to optimize the performance of the Supply Chain from the lens of multiple related disciplines including: Sales Forecasting, Warehousing/Inventory Management, Promotion, Pricing, Logistics Network Optimization, Freight Cost Management, Manufacturing, Retail POS Information, Ecommerce, Consumer Data, and Product Design/Packaging. After completing this course, you will be prepared to work in any of the numerous specialty areas possible in the world of Supply Chain Management.

Prerequisites:

• MSCA 31008: Data Mining Principles

• MSCA 37014: Python for Analytics

At the same time, Big Data and data analytics have raised numerous ethical questions about how data is collected, how data analytics methods transform that data into predictive information, and how that information is then used. This 3-hour seminar will introduce you to the benefits of Big Data and data analytics as well as the ethical challenges they raise. We will examine the rise of Big Data and data analytics as a fundamental paradigm shift in human inquiry; review ethical frameworks that help contextualize conversations about the benefits and challenges of Big Data and data analytics; study challenges related to privacy, anonymity, bias, and disparate impact; compare regulatory environments in the U.S., European Union, India, and China; and study how one large corporation has wrestled with its own Big Data and data analytics ethical challenges.

It covers the newest topics appearing in the curriculum of MScA at University of Chicago as well as foresights of invited leading industry specialists about main trends in the field of analytics. The students attend the course remotely. Activities include live presentations, workshops, individual and group projects and prerecorded videos for asynchronous learning. Upon completion of the course students receive a document certifying completion.

Prerequisites

• MSCA 31009: Machine Learning & Predictive Analytics (Required)
• MSCA 31013: Big Data Platforms (Recommended)
• MSCA 32017: Advanced Machine Learning and Artificial Intelligence (Recommended)

Prerequisite:

• Restricted to MSCA & MSAP students, and MScA Alumni Scholars only.