First Prediction : Baseline Prediction
First Model : Baseline Model
A.From a Data Perspective
-
Could we Collect more Data ? (Generally More Data, More Training, More Patterns, Better Accuracy and Better Model)
-
Could we Improve our Data ? (Data Cleaning, Adding New Feature or Remove Unnecessary Feature : Dimensionality Reduction)
B.From a Model Perspective
-
Is there any Better Model we can use for Training our Data ? (Try Estimators Like Random Forest Classifier)
-
Could we Improve the Current Model ? (Improving Parameters and Adjusting Hyperparameters)
Parameters : Model Find this Patterns in Data. (By Default Parameters Value)
Hyperparameters : Setting Parameter Values we can Adjust to Improve its Ability to Find Patterns.
A.Adjust Manually
B.Random Search Cross Validation
C.Grid Search Cross Validation
