To acquaint students with various concepts and technologies of data mining, with an emphasis on the possibility of multivariate statistical analysis and the use of software for its use in planning and business decision-making processes.

Develop an understanding of the capabilities and limitations of popular data analysis technologies.

Introduce technologies called "data mining" that help researchers recognize patterns and intelligently use the huge amounts of data collected through the Internet, electronic stores, banking, social networks, etc.

Knowledge and skills

• know key methods for classifying, forecasting, and describing data;
• be able to apply data analysis methods to real data;
• be able to evaluate and explain the results of various data analysis algorithms;
• be able to formulate data analysis tasks, identify appropriate technologies and implement them in relevant scientific and business projects;
• be able to identify possible applications of data analysis and business forecasting in business and scientific applications;
• use correctly, if necessary, in presentations and reports the vocabulary and terms of this subject.

Topics

1. Introduction. Overview of the main tasks and technologies of data analysis. Data mining process (CRISP-DM).
2. Data. Data types. Data pre-processing.
3. Descriptive statistics and data visualization.
4. Practical lesson 1. Visualization of multidimensional data.
5. Verification of statistical hypotheses and use in data analysis (marketing, reliability analysis of transport services, etc.).
6. Practical lesson 2. Analysis of data on cargo delays, public transport.
7. Forecasting methods. Multivariate regression analysis. Time series.
8. Practical lesson 3. Regression model for forecasting the volume of transportation of goods and passengers.
9. Case study 1. Short-term forecasting models for transportation.
10. Classification algorithms without training. Similarity measures of objects. Cluster analysis. Hierarchical clustering. K-means procedures. Clustering Validation.
11. Practical lesson 4. Segmentation of clients based on cluster analysis.
12. Classification algorithms without training. Discriminant analysis as a forecasting algorithm. Fraud detection, scoring.
13. Practical lesson 5. Credit rating.
14. The problem of reducing dimensionality and various ways to solve it. Principal component method. Factor analysis. Benchmarking and designing composite indicators.
15. Case study 2. Application of Data Mining methods in marketing and customer relationship management.
16. New technologies, methods and applications. Big data.