Please enable JavaScript.

Coggle requires JavaScript to display documents.

DATA MINING - Coggle Diagram

- - - - Process: Input data > Data processing > Data mining > Postprocessing > Information
  - - - Regression
      - Classification
- - - - Quantitative
        
        Nominal
        
        Ordinal
      - Numeric
        
        Interval
        
        Ratio
  - - - Noise
      - Artifact
      - Bias
      - Precision
      - Accuracy
    - - Outliners
      - Missing & Inconsistent
      - Duplicate
- - - - Induction
    - - Deduction
  - - - TNR
        
        Specificity, Selectivity
    - - FPR
        
        Type I Error, False Alarm, Fall Out
    - - FNR
        
        Type II Error, Miss Detection
    - - TPR
        
        Recall, Sensitivity, Hit rate
  - - - F1-Score
      - Macro Avg
      - Weighted Avg
    - - Underfitting
      - Overfitting
        
        Limited training size
        
        High model complexity
    - - Adjust Parameters
        
        Grid search
        
        Random search
      - Ensemble Methods
- - - - Class Distribution
      - Imbalance degree
- - - - Distribution
      - Many distribution at once
      - Other options for distribution
  - - - legend
      - No legend
- - - - Binary representation
    - - Association rule
  - - - Support determines how often a rule is applicable to a given data set.
    - - Confidence determines how frequently items in Y appear in transactions that contain X
  - - - Frequent Itemset Generation
      - Rule Generation
    - - Reduce the number of candidate itemsets
        
        The Apriori principle
        
        Using more advanced data structures,
    - - The Apriori Principle
      - Support-based pruning
      - Frequent itemset generation using the Apriori algorithm
- - - - Cluster analysis groups data objects based on information found only in the data that describes the objects and their
        relationships.
      - Segmentation
      - Partitioning
    - - Well-separated clusters
      - Center-based clusters
      - Contiguity-based clusters
      - Density-based clusters
      - Conceptual clusters
  - - - The goal is to find objects that do not conform to normal patterns or behavior.
      - Fraud Detection
      - Intrusion Detection
      - Ecosystem Disturbances
      - Medicine and Public Health
      - Aviation Safety
    - - Statistical approaches
      - Proximity-based Approaches
      - One-class Classification
      - Information Theoretic Approaches