Classification analysis

Classification analysis

MA609 Business Analytics and Data Intelligence Week 10: TUT 10

1. How businesses collect and produce data? They do it via:
   1. Sales and returns transactions
   1. Bar code scans
   1. Credit card transactions
   1. GPS and RFID tracking
   1. Clicks on a webpage

• Define data mining
• Data mining is the process of finding and extracting useful information and insights from large datasets

• Like geological mining
  • It is often hard, dirty work
  • It takes the right tools

• Explain data mining process
• Identify Opportunity
  • Don’t dig randomly
  • Begin with the end in mind
  • What is the business problem/opportunity?
• Collect Data
  • Decided where to dig
  • Get the right data – internally or externally
  • Millions of records aren’t required – use samples
  • 10p to 15p records is OK (where p = # of variables)

• Understand, Explore & Prepare the Data
  • Know what the data represents
  • Make sure it is clean & complete
  • Eliminate unneeded/redundant variables
  • Transform variables as needed
  • You might spend most of your data mining time here!

• Identify Task & Tools
  • Classification (supervised)
  • Prediction (supervised)
  • Segmentation/Clustering (unsupervised)
• Partition Data
  • Training
  • Validation
  • Testing (optional)
• Build & Evaluate Models
  • Try different models
  • Try different parameter settings
  • Avoid overfitting
• Deploy Models
  • Integrate models in operational systems
  • Train users
  • Monitor results
  • Look for opportunities for continuous improvement

• Define classification and give a few examples of its application.

Classification determines into which of m mutually exclusive group does an observation of unknown origin belong. Some areas of classification application are:

• Character/target recognition
• Oil/gold exploration
• Loan approval
• Diagnose diseases
• Identify defects
• Predict bond ratings
• Fraud detection (credit card, tax, trading, etc)
• Predict winners of sports events

• What are steps to classify using Full Bayes Classifier? What could be the problem in this case?
• To classify a new record
  • Find all matching records
  • Put new record in most frequently occurring matching group
• Problem
  • Continuous variables are unlikely to match exactly
  • Even with nominal variables, there might not be a match

The remainder of today’s session is allocated to the group project. When finished tutorial please work on the project.