Computational analysis for socio-economic sciences

Abstract

In this thesis we show three projects in the field of the physics of complex systems. All of them are based in the statistical analysis of real data, subfield nowadays commonly known as Data Science. Indeed, the last years have been characterized by a rapid growth of the amount of data. These data range from economy to biology, from finance to astrophysics, and many others. Generally the extraction of the information from them is a complex problem and it requires advanced statistical and mathematical frameworks. Here, we show three projects which are based in the statistical analysis of real data. In the first, entitled The light pollution as a surrogate for urban population of the US cities, we approached the problem of the light pollution analyzing the scaling of the population of the US cities with the night-time light. In the second project, entitled Dynamics in the Fitness-Income plane: Brazilian States VS World Countries we provide a variant of the Fitness algorithm (a novel method to compare the economic development of world countries) to measure the development of the Brazilian states. In the third and last project, entitled Dynamics of racial segregation and gentrification in New York City we focused in the analysis of the racial residential segregation. In that project we introduce a new index of segregation and we compare the racial residential segregation with several other factors such as the per capita income, the properties values, and flux of people, finding connections with the gentrification of some neighbors of New York City. We conclude the thesis providing the main results of each project and emphasizing the importance of the interaction among scientists of different areas in the study of socio-economic sciences.