2/20/26

• New paper, uses feature selection and feature scaling
• Feature selection helps model perform better + lowers the cost
• Uses a binary classification tabular dataset, fit the entire set of features
• Some algorithm to select subset of features
• They also do PCA w/ reducing the dimension
• Use selected features in ML models
• Read the paper, understand the paper, and ask questions
  • Try to find a dataset, only requirements is that project should not be done before
• For each dataset, let’s create a table with # of features and a categorial response variable
• Considerable amount of features (40-50)
• Maybe take like a couple weeks, if I get something, meeting next Friday at 10:30

2/27/26

• Try to find a response variable
• Then do feature selection and feature scaling
• Meet next Friday + try to finalize the dataset

3/6/26

• Get the entire dataset in one place
  • one csv file
• Response variable can be like levels using the rates

3/27/26

• Scale data that is not a percentage