Professional Documents
Culture Documents
Workflow For A New Dataset in Kaggle
Workflow For A New Dataset in Kaggle
9. Do Feature Engineering
● Use some grouping and categorization methods to group and
categories some data
● Create bins and group them under some category
● Again make plots with the dependent variable to find new
analysis
● Find out the type of skewness in the features and use appropriate
methods to reduce skewness and fill in missing values
● Use proper transformation techniques to engineer the features
● Also, convert categorical data to numerical data
● Make sure there is no difference in the training set and test set
10. Modelling
● Import necessary Classifiers classes
● Split the dataset into Features and Dependent Variable
● Apply Feature Scaling
● Use Different Models with Hyperparameter Tuning
● Choose the best model