The data mining process

The data mining process comprises different steps such as building, testing, or working with the mining models.

You begin a data mining project with a well-defined business intelligence project plan. The business analysts in your company define a problem that they want to solve, and a definite business intelligence goal that they want to achieve. The better this initial formulation, the clearer your guidelines about what data and which mining functions you use to achieve the desired result.

Your data mining project consists of the following major phases:
  1. Data selection and preparation.
  2. Building the data mining model (also known as the training phase).

    You build a data mining model from a specific set of input data.
    During the model-building process, after preparing your data you specify your decisions about:
    • Where the input data resides
    • Which fields in the input data are relevant
    • Which settings to use for the particular mining function that you are using
    • Where you want to store the final model
    Intelligent Miner® uses the following database objects when building a model:
    • User-defined methods
    • Stored procedures
    • User-defined functions
    .
  3. Testing a model and analyzing its quality.

    You can test a Classification or Regression model. Then you can analyze the quality of the model.

  4. Working with a model provides information on:
    • Visualizing the results.

      You can display your data mining results to analyze and interpret them. Use Intelligent Miner Visualizer to view and analyze the results.

    • Scoring data records.

      You apply a model to other data in the application phase of data mining. Use Intelligent Miner to score the data records.

    • Analyzing a model and preparing it for further processing steps.

      You can use various functions to retrieve information about the model in tables for further processing by other application programs.