Statistical Learning for business decision making: from big data to informative data

Abstract

In this thesis, we will pitch out these four aspects that are fundamental in modern businesses. The first aspect is about the growing importance of Statistics in business, which has two purposes: synthesizing and generalizing. Synthesizing means preparing the data collected in form (tables, graphs, numerical summaries) that enables a better understanding of the phenomena for which the survey was carried out. Synthesis meets the need for simplification, which stems from the human mind's limited capacity to handle articulated, complex, or multidimensional information. Descriptive Statistics is a specific field of Statistics that describe how to use techniques that allow a comprehensive study of a large amount of quantitative and qualitative information to highlight its characteristics, links, differences or associations among the variables surveyed. Generalizing means extending the result of the analysis performed on a limited group of statistical units (sample) data to the entire community to which it belongs (universe, population). This operation of generalizing is carried out according to methods of induction, which represent the content of Inferential Statistics. With the Inferential Statistics and Decision Theory closely related to the calculus of probability, it is possible to understand and interpret uncertain or random events and suggest rational rules of behavior and decision-making. .. [edited by Author]