Exploratory Data Analysis Task

An Exploratory Data Analysis Task is an exploratory study task that is a data analysis task that can formulate hypotheses through data pattern discovery.

• AKA: Exploratory Data Analysis, EDA Task, Data Exploration Task, Exploratory Analysis Task.
• Context:
  • It can typically discover Exploratory Data Patterns through exploratory visualization techniques and exploratory statistical methods.
  • It can typically identify Exploratory Data Structures via exploratory clustering algorithms and exploratory dimensionality reductions.
  • It can typically reveal Exploratory Data Anomalies through exploratory outlier detection and exploratory deviation analysis.
  • It can typically generate Exploratory Data Hypotheses via exploratory pattern recognition and exploratory relationship discovery.
  • It can typically assess Exploratory Data Quality through exploratory completeness checks and exploratory consistency evaluations.
  • ...
  • It can often uncover Exploratory Data Relationships via exploratory correlation analysis and exploratory association mining.
  • It can often detect Exploratory Data Distributions through exploratory histogram generation and exploratory density estimation.
  • It can often extract Exploratory Data Features via exploratory factor analysis and exploratory component analysis.
  • It can often reveal Exploratory Data Trends through exploratory time series visualization and exploratory pattern tracking.
  • It can often be triggered by an Exploratory Data Analysis Request.
  • ...
  • It can range from being a Visual Exploratory Data Analysis Task to being a Quantitative Exploratory Data Analysis Task, depending on its exploratory analysis approach.
  • It can range from being a Simple Exploratory Data Analysis Task to being a Complex Exploratory Data Analysis Task, depending on its exploratory analysis dimensionality.
  • It can range from being a Manual Exploratory Data Analysis Task to being an Automated Exploratory Data Analysis Task, depending on its exploratory analysis execution mode.
  • It can range from being a Single-Variable Exploratory Data Analysis Task to being a Multi-Variable Exploratory Data Analysis Task, depending on its exploratory analysis variable scope.
  • It can range from being a Structured Exploratory Data Analysis Task to being an Unstructured Exploratory Data Analysis Task, depending on its exploratory data format.
  • ...
  • It can be performed during an Exploratory Analysis Phase as an Exploratory Data Analysis Activity.
  • It can precede a Confirmatory Data Analysis Task in data analysis workflows.
  • It can be supported by Exploratory Visualization Systems implementing exploratory visualization algorithms.
  • It can integrate with Interactive Analysis Platforms for exploratory data manipulation.
  • It can leverage Statistical Computing Environments for exploratory computations.
  • It can connect to Data Profiling Tools for exploratory metadata generation.
  • ...
• Example(s):
  • Exploratory Clustering Tasks, such as:
    • Exploratory Hierarchical Clustering Tasks, such as:
      • Exploratory Dendrogram Analysis for exploratory cluster hierarchy visualization.
      • Exploratory Agglomerative Clustering for exploratory group formation.
      • Exploratory Divisive Clustering for exploratory subset identification.
    • Exploratory Partitioning Tasks, such as:
      • Exploratory K-means Analysis for exploratory centroid discovery.
      • Exploratory DBSCAN Analysis for exploratory density-based grouping.
      • Exploratory Spectral Clustering for exploratory graph-based partitioning.
  • Exploratory Association Tasks, such as:
    • Association Rule Mining Tasks, such as:
      • Market Basket Analysis Task for exploratory purchase pattern discovery.
      • Frequent Itemset Mining Task for exploratory co-occurrence identification.
      • Sequential Pattern Mining Task for exploratory temporal association discovery.
    • Exploratory Correlation Tasks, such as:
      • Exploratory Pearson Correlation Analysis for exploratory linear relationship discovery.
      • Exploratory Spearman Correlation Analysis for exploratory rank relationship identification.
      • Exploratory Mutual Information Analysis for exploratory nonlinear dependency detection.
  • Exploratory Visualization Tasks, such as:
    • Exploratory Statistical Visualization Tasks, such as:
      • Box Plot Analysis Task for exploratory distribution comparison.
      • Scatter Plot Analysis Task for exploratory bivariate relationship visualization.
      • Histogram Analysis Task for exploratory frequency distribution display.
    • Exploratory Multivariate Visualization Tasks, such as:
      • Parallel Coordinates Analysis for exploratory high-dimensional pattern discovery.
      • Heatmap Analysis Task for exploratory matrix pattern visualization.
      • Multi-vari Chart Analysis for exploratory variation source identification.
  • Exploratory Dimensionality Reduction Tasks, such as:
    • Exploratory Factor Analysis Tasks, such as:
      • Exploratory Principal Component Analysis for exploratory variance maximization.
      • Exploratory Independent Component Analysis for exploratory signal separation.
      • Exploratory Non-negative Matrix Factorization for exploratory parts-based representation.
    • Exploratory Manifold Learning Tasks, such as:
      • Exploratory t-SNE Analysis for exploratory neighborhood preservation.
      • Exploratory UMAP Analysis for exploratory topological structure preservation.
      • Exploratory Isomap Analysis for exploratory geodesic distance preservation.
  • Exploratory Statistical Analysis Tasks, such as:
    • Exploratory Descriptive Statistics Tasks, such as:
      • Five-Number Summary Analysis for exploratory distribution characterization.
      • Letter Value Display Analysis for exploratory tail behavior examination.
      • Stem-and-Leaf Plot Analysis for exploratory data shape visualization.
    • Exploratory Robust Statistics Tasks, such as:
      • Median Polish Analysis for exploratory two-way table decomposition.
      • Resistant Line Analysis for exploratory outlier-resistant regression.
      • Trimean Analysis for exploratory robust central tendency estimation.
  • Domain-Specific Exploratory Analysis Tasks, such as:
    • Exploratory Text Analysis Tasks, such as:
      • Topic Discovery Task for exploratory theme identification.
      • Word Cloud Analysis Task for exploratory term frequency visualization.
      • Text Clustering Task for exploratory document grouping.
    • Exploratory Network Analysis Tasks, such as:
      • Community Detection Task for exploratory network structure discovery.
      • Centrality Analysis Task for exploratory node importance identification.
      • Network Visualization Task for exploratory relationship pattern display.
  • Interactive Exploratory Analysis Tasks, such as:
    • OLAP Analysis Tasks (Online Analytical Processing), such as:
      • Drill-Down Analysis Task for exploratory detail examination.
      • Roll-Up Analysis Task for exploratory aggregation exploration.
      • Slice-and-Dice Analysis Task for exploratory dimensional filtering.
    • Dashboard-Based Exploration Tasks, such as:
      • Interactive Filtering Task for exploratory subset investigation.
      • Dynamic Visualization Task for exploratory real-time pattern discovery.
      • What-If Analysis Task for exploratory scenario exploration.
  • ...
• Counter-Example(s):
  • Confirmatory Data Analysis Task, which tests predefined hypotheses rather than generates exploratory hypotheses.
  • Predictive Modeling Task, which creates predictive models rather than explores data patterns.
  • Data Processing Task, which transforms data formats rather than discovers data insights.
  • Statistical Hypothesis Testing Task, which validates specific claims rather than explores open-ended questions.
  • Operational Reporting Task, which produces standard reports rather than discovers unexpected patterns.
• See: Knowledge Discovery Task, Exploratory Study Task, Abductive Reasoning, John W. Tukey, Data Visualization Task, Unsupervised Learning Task, Pattern Recognition Task.

---

---

References

2011

• (Wikipedia, 2011) ⇒ http://en.wikipedia.org/wiki/Exploratory_data_analysis
  • QUOTE: In statistics, exploratory data analysis (EDA) is an approach to analysing data sets to summarize their main characteristics in easy-to-understand form, often with visual graphs, without using a statistical model or having formulated a hypothesis. Exploratory data analysis was promoted by John Tukey to encourage statisticians visually to examine their data sets, to formulate hypotheses that could be tested on new data-sets. …

    … Typical graphical techniques used in EDA are: Box plot; Histogram; Multi-vari chart; Run chart; Pareto chart; Scatter plot; Stem-and-leaf plot; Odds ratio; Chi-square; Multidimensional scaling; and Targeted Projection Pursuit.

    Typical quantitative techniques are: Median polish; the Trimean; Letter values; Resistant line; Resistant smooth; Rootogram; Ordination

2009

• http://www.itl.nist.gov/div898/handbook/eda/eda.htm
• http://www.itl.nist.gov/div898/handbook/eda/section1/eda11.htm
  • Exploratory Data Analysis (EDA) is an approach/philosophy for data analysis that employs a variety of techniques (mostly graphical) to
    • 1. maximize insight into a data set;
    • 2. uncover underlying structure;
    • 3. extract important variables;
    • 4. detect outliers and anomalies;
    • 5. test underlying assumptions;
    • 6. develop parsimonious models; and
    • 7. determine optimal factor settings.
  • Focus: The EDA approach is precisely that--an approach--not a set of techniques, but an attitude/philosophy about how a data analysis should be carried out.

2000

• (Anselin et al., 2000) ⇒ Luc Anselin, Jacqueline Cohen, David Cook, Wilpen Gorr, and George Tita. (2000). “Spatial Analyses of Crime.” In: Criminal justice, 4(2).
  • QUOTE: … Special attention is given to some practical and accessible methods of exploratory data analysis that arguably should be the starting place of any empirical analyses of the relationship of place to crime.

1999

• (Zaiane, 1999) ⇒ Osmar Zaiane. (1999). “Glossary of Data Mining Terms." University of Alberta, Computing Science CMPUT-690: Principles of Knowledge Discovery in Databases.
  • QUOTE: Exploratory Data Analysis: The use of graphical and descriptive statistical techniques to learn about the structure of a dataset.

1977

• (Tukey, 1977) ⇒ John W. Tukey. (1977). “Exploratory Data Analysis. Addison-Wesley. ISBN:0201076160

---