Scikit-Learn

Attributes 
Followings table consist the attributes used by 
Logistic Regression
module: 
Attributes 
Description 

---

Scikit-Learn
37 
coef_
: 
array, 
shape(n_features,) 
or 
(n_classes, n_features) 
It is used to estimate the coefficients of the 
features in the decision function. When the given 
problem is binary, it is of the shape (1, 
n_features). 
Intercept_
: array, shape(1) or (n_classes) 
It represents the constant, also known as bias, 
added to the decision function. 
classes_
: array, shape(n_classes)
It will provide a list of class labels known to the 
classifier. 
n_iter_
: array, shape (n_classes) or (1) 
It returns the actual number of iterations for all the 
classes.
Implementation Example 
Following Python script provides a simple example of implementing logistic regression on 
iris
dataset of scikit-learn: 
from sklearn import datasets 
from sklearn import linear
_
model 
from sklearn.datasets import load
_
iris 
X, y = load_iris(return_X_y=True) 
LRG = linear
_
model.LogisticRegression(random_state=0,solver='liblinear',multi 
).fit(X, y) 
LRG.score(X, y) 
Output 
0.96 
The output shows that the above Logistic Regression model gave the accuracy of 96 
percent. 
Ridge Regression 
Ridge regression or Tikhonov regularization is the regularization technique that performs 
L2 regularization. It modifies the loss function by adding the penalty (shrinkage quantity) 
equivalent to the square of the magnitude of coefficients.
∑ (𝑌
𝑖
− 𝑊
0
− ∑ 𝑊
𝑖
𝑋
𝑗𝑖
𝑛
𝑖=1
)
2
+ 
𝑚
𝑗=1
𝛼 ∑ 𝑊
𝑖
2
𝑛
𝑖=1
= 𝑙𝑜𝑠𝑠_𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 + 𝛼 ∑ 𝑊
𝑖
2
𝑛
𝑖=1
