Please enable JavaScript.

Coggle requires JavaScript to display documents.

DATA SCIENCE (CLASSIFICATION (DECISION TREE (GENERAL (USE (PREDICTION,…

- - - - USE
        
        PREDICTION, BINARY / NOMINAL CATEGORIES
        
        LESS EFFECTIVE FOR ORDINAL CATEGORIES
        
        EXPLORATORY, DISTINGUISH OBJECTS BTW CLASSES
      - INPUT
        
        COLLECN. RECORDS (INSTANCES)
        
        TUPLE(X,Y)
        
        X: ATTRIBUTE SET
        
        CONTINUOUS
        
        DISCRETE
        
        Y: CLASS LABEL, TARGET ATTRIBUTE
        
        MUST DISCRETE
        
        DISTINGUISH W REGRESSION
      - ASSIGN OBJECTS TO 1 PREDEFINED CATEGORIES
      - KEY OBJECTIVE: GOOD GENERALIZAN. CAPABILITY
        
        ACCURATE PREDICN. CLASS LABELS OF PREVIOUSLY UNKNOWN RECORDS
      - SUPERVISED STRATEGY
        
        SPLIT 70 TRAINING /30 TESTING
        
        IN DEPLOYMENT ENV. : POOR PERF
        
        DEV SET 20%: TUNE PARAMETERS, SELECT FEATURES
        
        TEST SET 10%:
        
        70% TRAINING