Please enable JavaScript.

Coggle requires JavaScript to display documents.

Chapters 13-15 (Chapter 13 (data robot (will accept CSV (sometimes better…

- - - - sometimes better to use TSV
        
        text data is less likely to be problematic
    - - .zip, .tar, .gz, .tgz, tar.bz2
- - - - convert ? to "missing data"
    - - only numeric data
    - - bracket= inclusive, parenthesis = exclusive
      - (1,3)=2, (1,3]=2,3
    - - multi-class
      - binary
      - numeric
      - text
      - boolean
- - - - "readmitted" in the prediction bar
    - - metric optimization
      - partitioning method
        
        partition: split into parts
        
        random and stratified give same options
      - holdout percentage
      - if intent is to run cross validation on the data, each additional fold leads to one more run-through of creating models
        
        5 and 10 folds are most common