• Print

Data Mining Checklist

 
Checklist to be used as a sequential guide to automatically explore large amounts of data to uncover patterns.
 
  • Mark all that apply.
      •  
    • GOAL DEFINITION - What is the goal/objective of your data mining project?
    • DATA SELECTION - What is the data needed for the project and what is/are the sources(s) for this data?
    • DATA PREPARATION - What processing does the data need to undergo before it is ready to be mined? Does it need to be sorted? Do any exclusion criteria need to be applied? Do multiple data sources need to be aggregated?
    • DATA EXPLORATION - Have you undergone any data verification processes to validate the data set prior to pattern discovery?
    • PATTERN DISCOVERY - Have you determined a pattern discovery algorithm and applied it to the data set?
    • PATTERN DEPLOYMENT - Have you applied the patterns discovered in the data to the initial business goal/objective of your data mining project?
    • PATTERN PRESENTATION - Have you communicated the pattern and its relationship to the business goal/objective to relevant stakeholders?
    • BUSINESS INTELLIGENCE - Have you deployed the discovered patterns as queries against a database to derive reports?
    • DATA SCORING AND LABELLING - Have you applied the discovered patterns to score and label each data record in the database?
    • DECISION SUPPORT SYSTEMS - Have the discovered patterns been used to make a decision support system?
    • ALARM MONITORING - Have the discovered patterns been established as norms for a business process against which deviations from the normal pattern can be set up with alarms?
    • PATTERN VALIDITY MONITORING - As the business process changes over time, has the validity of the discovered patterns been re-evaluated?
     
    Pyzdek, Thomas & Keller, Paul A. "The Six Sigma Handbook" 3rd Edition. McGraw Hill (2010), p. 110-112.