Exploratory Data Analysis

Exploratory Data Analysis (EDA) is an approach employed to analyze datasets. Primarily, EDA uses data visualization methods and often statistical models to (i) assess a dataset’s general structure, (ii) obtain descriptive summaries of the data, and (iii) provide the basis for model formulation. EDA includes checks on data quality, calculation of summary statistics, and data plotting. The data quality is checked regarding errors, outliers, and missing observations. This chapter explores simple visualization EDA techniques, algorithms to detect and handle outliers and missing values, more advanced tools such as correlograms and clustering, and dimensionality reduction techniques.