Regularization with logistic regression
If you have already worked your way through Chapter 7, Linear Regression Models, and read the first section of this chapter, you already have a good idea of how regularization works. We add a penalty to the estimator that minimizes our parameter estimates. The size of that penalty is typically tuned based on a measure of model performance. We will work through that in this section. Follow these steps:
- We will load the same modules that we worked with in the previous section, plus the modules we will need for the necessary hyperparameter tuning. We will use
RandomizedSearchCVanduniformto find the best value for our penalty strength:import pandas as pd import numpy as np from sklearn.model_selection import train_test_split from sklearn.preprocessing import StandardScaler from sklearn.preprocessing import OneHotEncoder from sklearn.pipeline import make_pipeline from sklearn.impute import SimpleImputer from sklearn.compose import ColumnTransformer...