Business Intelligence Chapter 2: Descriptive Analytics

Business Intelligence, Analytics, and Data Science: A Managerial Perspective - Chapter 2

• Descriptive Analytics I: This chapter covers the nature of data, statistical modeling, and visualization.

• Data: A collection of facts; usually obtained from experiences, observations, or experiments. Data can include numbers, words, images, etc.

• Data Quality and Integrity: Critical to analytics. Data integrity includes accuracy, completeness, consistency, and validity of an organization's data.

• Data Types:

  • Structured Data: Standardized format, well-defined structure, complies with a data model, and follows a persistent order, easily accessed. Examples: names, dates, addresses, credit card numbers, stock info, geolocation.
  • Unstructured Data: Any combination of textual, imagery, voice, and web content.
  • Semi-structured Data: Extensible markup language (XML), hypertext markup language (HTML), JavaScript Object Notation (JSON), log files, etc.

• Metrics for Analytics Ready Data:

  • Data source reliability
  • Data content accuracy
  • Data accessibility
  • Data security and data privacy
  • Data richness
  • Data consistency
  • Data currency/data timeliness
  • Data validity and relevance
  • Data granularity

• Categorical Variables:

  • Nominal Data: Used to label variables without quantitative value. Examples: gender, hair color, nationality, marital status.
  • Ordinal Data: Variables ordered based on their relative position. The differences between the values aren't necessarily consistent or meaningful (e.g., Likert scales: very likely, likely, neutral, unlikely, very unlikely).

• Numerical Variables:

  • Interval Data: Variables with order and difference. Examples: Classification of people (teenager, youth, middle-age, etc.)
  • Ratio Data: Order and difference between variables, with a true zero point. Examples: income, height, weight, sales, etc.

• Data Preprocessing: The process of preparing data for analytics. This includes data consolidation, data cleaning, data transformation, and data reduction.

• Data Reduction Techniques:

  • Variables: Dimensional reduction, variable selection
  • Cases/samples: Sampling, balancing/stratification, discretization

• Dimensionality Reduction: A process to reduce the number of dimensions (features) in a dataset while preserving the most important properties. This is useful for large, high-resolution images.

• Discretization: Converting continuous data into discrete intervals or categories. Typically applied to large datasets to simplify analysis or model building (e.g., age groups, income brackets).

• Data Normalization Reorganizing data to remove unstructured or redundant information, enabling a standardized data format. This creates an organized data system.

• Data Visualization Using visual representations for exploring, understanding and communicating data. Often incorporates charts, graphs, illustrations.

• Statistical Modeling for Business Analytics: Uses statistics to analyze data. Includes:

  • Descriptive Statistics: Describing data (e.g., mean, median, mode, range, variance).
  • Inferential Statistics: Drawing conclusions about a population based on a sample (e.g., hypothesis testing, regression, various graphs/charts).

• Regression Modeling: A statistical technique to characterize the relationship between explanatory and response variables. Used for hypothesis testing and forecasting.

• Business Reporting:

  • Report: Information to drive decisions, acting as a communication artifact.
  • Report Composition: Includes sources, format (text, table, graphs), distribution channels (in-print, email, portal).
  • Dashboard Design Considerations: The design should present data in a clear and concise manner so information is easily accessible, enabling rapid assimilation.

• Performance Dashboards: Used in Business Process Management (BPM) and Business Intelligence (BI) platforms. Provide visual displays of information, combined on a single screen for quick analysis. These dashboards facilitate drilling into data and exploration.