Nowadays, in the period of the digitalization and knowledge economy development, majority of activities result in the increase of data that should be captured. In each area of business, there is an increasing urge to extract the knowledge from data in a timely manner in order to be able to make shifts that ensure a competitive advantage. Thus, the knowledge of methods and techniques of big data processing is currently a tool of a modern economist. Courses oriented to how to use machine learning tools in the economic practice should become a part of a study plan at universities of economics and business. In our paper, we propose the concept of teaching of machine learning techniques using software Weka as a possible solution to integrate this issue into the study plan based on the example of problem-based learning.