Data science: um glossário para profissionais da área de química

Abstract

Data Science aims to provide strategies and tools capable of optimizing processes and increasing results strength, being a powerful and decisive mechanism to enhance the research and production capability, which ensures the conformity of technology centers and laboratories. The statistical methods comprise the basis which underpin data science, in its varied applications and distinctions, being invaluable for researchers of all sectors. Several tools and computing languages stand out as they aim to enable or facilitate the data analysis. Those tool work together with the Design of Experiments, being fundamental for obtain conclusive data and making assertive decisions. The optimization processes, along with cost and waste reduction, lead to standardization of the production center and allow the interface between Data Science, Design of Experiments and quality control processes. The programming tools for Data Science play a roll since the beginning of the exploratory analysis of the data project, starting with the reading and comprehension of text files (txt), CSV (comma-separated values), XML (extensible markup language) and others. These obtained data are treated and modeled, becoming appropriate to go through mathematical and statistical methods, obtain specific information for conclusions about experiments, based on descriptive and inferential statistics, hypothesis tests, ANOVA and linear regression. This work is centered on the description of the following programming tools and quality control methods: R, Python, C++, Matlab, Octave, SQL, 6σ and 5S.