Key Components of Exploratory Data Analysis (EDA)
03/05/2024 | by Patrick Fischer, M.Sc., Founder & Data Scientist: FDS
1. Descriptive Statistics:
- Measures of central tendency: Calculation of means, medians, and modes.
- Measures of dispersion: Analysis of variability through standard deviation, quartiles, and range.
2. Visualization Techniques:
- Histograms, Boxplots, Scatterplots, Heatmaps, Pair Plots.
3. Univariate Analysis:
- Examination of a single variable.
4. Bivariate Analysis:
- Exploration of relationships between two variables.
5. Multivariate Analysis:
- Analysis of relationships involving more than two variables.
6. Identification of Outliers:
- Application of methods like IQR or Z-Score to identify outliers.
7. Imputation of Missing Data:
- Determination of strategies for handling missing data.
8. Data Transformation:
- Application of transformations such as logarithms, standardization, or normalization.
9. Hypothesis Generation:
- Formulation of hypotheses based on exploratory analysis.
10. Contextualization:
- Consideration of the context of the data and the domain.
Exploratory Data Analysis is an iterative and interactive process that lays the foundation for further statistical analysis and model building.