How to Think Like a Data Scientist: Application of a Variable Order Markov Model to Indicators Management

摘要

The growing demand for specialists in analyzing large volumes of data has led to an emerging profile of knowledge managers known as data scientists. How to address the different and complex scenarios with mathematical methods makes a difference when to apply them successfully in a dynamic environment such as the management indicators. For this reason, the authors present in this article a case study of prognostic indicators, developed in the field of finance, making use of mathematical Markov model which has prototyped in an abstract technological implementation with the capabilities to implement cases in other contexts. The purpose of the case study is to verify if the different levels of analysis of the Markov model provide knowledge to the prognosis by indicators while the application of the proposed methodology is shown. Thus, this work introduces to the threshold of a methodology that leads to one of the ways on how to think like a data scientist.