7 PM – AI/ML – Darin (Joel Sub)

Today we did:

1. We checked the progress of the final projects and continued working on the Titanic project.

Homework:

1. For your final projects, do as much as you can of the following: You’ll want to train 1-2 other models as well. Try researching other resnet backbones instead of just resnet34 which we always use. Train those models as well, and do some research to compare those models. Then, based on that comparison, export the best model and use that for streamlit, which at that point you can get started on as well.
2. In the Titanic notebook:
   #Create a variable called dep_var and set to the column of the target variable (“Survived”)
   dep_var = “Survived”

   #Create a list called category_names
   #List all the columns/features that are categorical
   #Create a list called continuous_names
   #List all the columns/feature that are continuous
   #Categorical: fits into categories. Example: sex, Pclass
   #Continuous: not categories but could be any number. Example: age, SibSp
   #Example list of category_names: category_names = [“Sex”] #Exclude name, ticket, and cabin from the lists