My Blog.

Phase 3 - Model Planning

Model Planning Overview: Model planning involves deciding on the methods, techniques, and workflow to be used in the subsequent model building phase. The team explores the data to understand relationships between variables and selects key variables and suitable models accordingly.

Key Activities:

  1. Assess Data Structure:

    • Evaluating the structure of the data helps determine the tools and analytic techniques needed for the next phase.

  2. Alignment with Objectives:

    • Ensuring that the chosen analytic techniques align with business objectives and can validate or invalidate working hypotheses.

  3. Model Complexity:

    • Deciding whether to employ a single model or a combination of techniques within a larger analytical workflow.

  4. Research:

    • Investigating how other analysts have tackled similar problems to gain insights and inform decision-making.

Data Exploration and Variable Selection:

  • Exploring data relationships through visualization tools.
  • Involving stakeholders and subject matter experts to incorporate their ideas and hypotheses.
  • Iteratively testing and refining to identify essential predictors and variables for the analysis.

Model Selection:

  • Choosing analytical techniques aligned with project goals.
  • Constructing models that mimic real-world behavior with sets of rules and conditions.
  • Deciding on techniques suited for structured, unstructured, or hybrid data.
  • Initial model creation using statistical software like R, SAS, or Matlab, with consideration for scalability to large datasets.

Common Tools:

  • R Programming Language: Offers comprehensive modeling capabilities with thousands of packages for data analysis and visualization.
  • SQL Analysis Services: Enables in-database analytics of common data mining functions and predictive models.
  • SAS/ACCESS: Facilitates integration between SAS and the analytics sandbox through various data connections.

In Simple Terms: Model planning is like deciding how to build a puzzle before putting the pieces together. First, you examine the puzzle's structure to understand how the pieces fit. Then, you make sure your plan matches the picture on the box (business objectives) and consider how complex your puzzle will be.

Next, you explore the pieces, trying to find the important ones and figuring out where they go. You might ask friends for advice (stakeholders) and test different arrangements until it looks right.

Once you know which pieces are crucial, you pick the right tools to help you assemble them. Whether you use a standard toolbox or a specialized one depends on the puzzle's size and complexity. And when you have everything ready, you're all set to start building your puzzle – or in this case, your analytical model!

MM - Phase 3 - Model PlanningMM - Phase 3 - Model PlanningFor organizing and enhancing your notes on Phase 3 - Model Planning of the Data Analytics Lifecycle, let's refine the structure and identify keywords or short phrases suitable for creating an effective mind map. This will help in better recall and understanding of the concepts. Refined Notes on Model Planning Overview Purpose**: Establish methods and workflow for model building. Data Understanding**: Analyze relationships and select key variables and models. Key Activities 1. Assess Data St