Please enable JavaScript.

Coggle requires JavaScript to display documents.

Data mining (Data and data type (Data : (Categorical (Ordinal, Nominal),…

- - - - Data matrix
      - Document data
      - Transaction data
  - - - Ordinal
      - Nominal
    - - Interval
      - Ratio
- - - - Noise
      - Outliers
      - Missing values
        
        Reasons:
        
        information is not collected or lost
        
        attributes may not be applicable to all cases
        
        Handling:
        
        remove
        
        Ignore
        
        Replace with static value
        
        replace with randome value
        https://developerzen.com/data-mining-handling-missing-values-the-database-bd2241882e72
        https://towardsdatascience.com/how-to-handle-missing-data-8646b18db0d4
        https://measuringu.com/handle-missing-data/
      - Duplicates
      - Domain specific proplems
- - - - Type of sampling:
        
        Simple Random Sampling
        
        Stratifield sampling
        
        Sampling without replacement
        
        Sampling with replacement
      - Sample zise: Note to take too small sample
    - - Techniques:
        
        Embedded approach
        
        Feature selection occurs naturally as part of the data
        mining algorithim
        
        Filter approach
        
        Features are selected before data
        mining algorithm is run
        
        Wrapper approach
        
        Use a data mining algorithm as a black box to
        find best subset of attributes
        
        Brute-force approach
        
        Try all possible feature subsets as input to data
        mining algorithim
    - - 3 general methodologies:
        
        Feature Extraction
        
        Feature construction
        
        Mapping data to new space
    - - Techniques:
        
        PCA(Priciple Component Analysis)
        
        SVD(Singular Value Decomposition)