2019/2020

Data Science for Business

Category 'Best Course for Broadening Horizons and Diversity of Knowledge and Skills'

Type: Mago-Lego

Delivered by: School of Data Analysis and Artificial Intelligence

When: 4 module

Instructors: Anvar Kurmukov, Leonid E Zhukov

Language: English

ECTS credits: 3

Contact hours: 40

Course Syllabus

Abstract

Data Science discipline formed recently in response to increasing use of data in business. It utilizes data mining, machine learning and statistical methods, but focuses on business applications, solving real world problems and delivering impact on business. This course is using a case-based approach to teaching, i.e. the students will be learning methods and techniques while solving business case problems. Each case will contain a description of a business problem and available data. The goal would be to convert a business problem into analytical and solve it using data with the help of variety of data mining and machine learning methods. The methods will be introduced as needed for each case solution. The course will be hands-on, during the lectures students will learn the approach and implement and solve the case in their home assignments.

Learning Objectives

• Providing students with essential knowledge of data mining methods and algorithms and experience in converting business problems into analytical and solving them.

Expected Learning Outcomes

• Students know basic notation and terminology used in data science.
• Students understand basic principles behind analysis algorithm.
• Students visualize, summarize and analyze datasets.
• Students formulate and solve analytical problems for given business problem.

Course Contents

• Introduction to Data Science for Business
  Introduction to a new discipline Data Science. Its place in academic world and industry. Examples of real world problems.
• Dealing with data
  Skills needed to work with data. Data cleaning and preparation. Basic data analysis.
• Data mining, machine learning, statistics.
  Major classes of algorithms, applicability, solution quality metrics.
• Case study 1. Customer segmentation
  The goal of the case is to group customers into clusters based on some customer similarity metrics. Algorithms: clustering – k-means, agglomerative, dimensionality reduction - PCA.
• Case study 2. Customer churn modeling
  The goal of the case is to predict which customers are going to leave the service within a given time. Algorithms: Supervised learning – logistic regression, decision trees, random forest.
• Case study 3. Pricing
  The goal of the case is to determine the optimal pricing for goods and services. Algorithms: supervised learning – regression (linear and non-linear models)
• Case study 4. Production optimization
  The goal of the case is to predict an output of the production line and find optimal parameter setting. Algorithms: supervised learning – regression, non-linear optimization.
• Case study 5. Sales territory design
  The goal of the case is to select locations of the sales offices to maximize the coverage under constrained resources. Algorithms: clustering and geo-analytics approaches.
• Dealing with big and fast data
  Handling data in real world – big data and data streams.
• Impacting the business
  How to create a visible impact on business with analytics

Assessment Elements

• Homework 1
• Homework 2
• Homework 3
• Homework 4
• Homework 5

Interim Assessment

• Interim assessment (4 module)
  0.2 * Homework 1 + 0.2 * Homework 2 + 0.2 * Homework 3 + 0.2 * Homework 4 + 0.2 * Homework 5

Bibliography

Recommended Core Bibliography

• James, G. et al. An introduction to statistical learning. – Springer, 2013. – 426 pp.
• Provost, Foster, Fawcett, Tom. Data Science for Business: What you need to know about data mining and data-analytic thinking. – " O'Reilly Media, Inc.", 2013.

Recommended Additional Bibliography

• Siegel, E. Predictive analytics: The power to predict who will click, buy, lie, or die. – John Wiley & Sons, 2016. – 338 pp.