Data Mining Process

Data Mining Process

1. Data Mining Task Types

   • Data Mining Tasks

2. 5 Stages

   • Data Pre-processing - Cleaning up the data
   • EDA - Investigate the data with visualisation (To decided what to do in the KDD)
   • Data Selection - Selecting what you want to mine
   • KDD - The mining part
   • Output - The results of the mining

3. Method 2 - Virtuous Cycle - Berry & Linoff 1997

   • 1. identifying the problem
   • 2. analysing the data
   • 3, Taking action
   • 4. Measuring the outcome
   • VERY SIMPLE

4. SEMMA - Cycle

   • Sample - Data Picking
   • Explore - EDA
   • Modify - Data Pre-processing
   • Model - KDD - Regression, Classification and neural networks
   • Assess - Graphs and Reports. Is the output useful?
   • FOCUS ON THE PROCESS OF DM

5. CRISP-DM

   • Includes the business objectives
     • How the data is deployed into the business

Imagens e arquivos do mapa