Changing coefficients
With this understanding of the coefficients of a LogisticRegression()
model, have a closer look at them to see how they change depending on what columns are used for training. Will the column coefficients change from model to model?
You should .fit()
two different LogisticRegression()
models on different groups of columns to check. You should also consider what the potential impact on the probability of default might be.
The data set cr_loan_clean
has already been loaded into the workspace along with the training sets X1_train
, X2_train
, and y_train
.
This exercise is part of the course
Credit Risk Modeling in Python
Exercise instructions
- Check the first five rows of both
X
training sets. - Train a logistic regression model, called
clf_logistic1
, with theX1
training set. - Train a logistic regression model, called
clf_logistic2
, with theX2
training set. - Print the coefficients for both logistic regression models.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Print the first five rows of each training set
print(____.____())
print(____.____())
# Create and train a model on the first training data
____ = ____(solver='lbfgs').____(____, np.ravel(y_train))
# Create and train a model on the second training data
____ = ____(solver='lbfgs').____(____, np.ravel(y_train))
# Print the coefficients of each model
print(____.____)
print(____.____)