Regression and Classification Metrics in Machine learning with Python

Model evaluation with metric API for regression and classification

In this article, we will discuss various metrics of regression and classification in machine learning. We always think of steps involved in modeling a good machine learning algorithm. The one-step is the metrics for evaluation of the goodness of the model. When we fit our model and make a prediction, then we always try to know the error and the accuracy. This article will try to deliver and explain various error measurement methods in regression and classification.

There are criteria to evaluate the prediction quality of the model as shown below:

• Metric functions: that we will study in this article.
• Estimator score method: this method has a score method to evaluate to solving the problem.
• Scoring parameter: The scoring parameters tells the estimator to choose the metric for evaluation of the model with grid_search.GridSearchCV and cross_validation.cross_val_score

Basic definition

• Estimator: It is a function or equation to predict the more accurate modeling points on real data points.

Tricks to know

There are two things to be noticed in the evaluation methods as shown below:

• First, some methods end with score word it means the value comes in this determined the ground truth. In this, if the number is on the higher side then it is better.
• The second, if the word ends with error or loss. In this, the number is lesser the better.

Metrics in Regression

The metrics for evaluate performance in regression are given below:

1. Explained Variance Score: This metric evaluates the variation or dispersion of the data points.

• The formula of this metric is shown below:

#import the variance score from the sklearn 
from sklearn.metrics import explained_variance_scoretrue_values = [5, 2.5, 3, 6]
predicted_values = [4.5, 2.9, 3, 7]explained_variance_score(true_values, predicted_values)#output:
0.8525190839694656

#import the max error from the sklearn
from sklearn.metrics import max_errortrue_values = [5, 2.5, 3, 6]
predicted_values = [4.5, 2.9, 3, 8]max_error(true_values, predicted_values)#output:
2

##import the MAE from the sklearn
from sklearn.metrics import mean_absolute_errortrue_values = [5, 2.5, 3, 6]
predicted_values = [4.5, 2.9, 3, 7]mean_absolute_error(true_values, predicted_values)#output:
0.475

#import the MSE from the sklearn
from sklearn.metrics import mean_squared_errortrue_values = [5, 2.5, 3, 6]
predicted_values = [4.5, 2.9, 3, 7]mean_squared_error(true_values, predicted_values)#output:
0.3525

#import the r-squared from the sklearn
from sklearn.metrics import r2_scoretrue_values = [5, 2.5, 3, 6]
predicted_values = [4.5, 2.9, 3, 7]r2_score(true_values, predicted_values)#output:
0.8277862595419847

#import the accuracy score from the sklearn
from sklearn.metrics import accuracy_scoretrue_values = [5, 2, 3, 6]
predicted_values = [4, 3, 3, 6]accuracy_score(true_values, predicted_values)

#import the classification report from the sklearn
from sklearn.metrics import classification_reporttrue_values = [3, 4, 3, 6]
predicted_values = [4, 3, 3, 6]
target_names = ['Apple', 'Orange', 'Kiwi']print(classification_report(true_values, predicted_values, target_names=target_names))

#importing the libraries
from sklearn import svm
from sklearn.metrics import hinge_loss
from sklearn.svm import LinearSVC#data set in x and y values
x_values = [[3], [2]]
y_values = [-1, 1]#using linear SVC model
svm_linear = svm.LinearSVC(random_state=0)#fitting the model
svm_linear.fit(x_values, y_values)
LinearSVC(random_state=0)#making decision prediction
pred_decision = svm_linear.decision_function([[-2], [3], [0.5]])
hinge_loss([-1, 1, 1], pred_decision)#output:
1.333372678152829