Stacking in Machine Learning
Stacking:
Stacking is a way of ensembling classification or regression models it consists of two-layer estimators. The first layer consists of all the baseline models that are used to predict the outputs on the test datasets. The second layer consists of Meta-Classifier or Regressor which takes all the predictions of baseline models as an input and generate new predictions.
Stacking Architecture:
mlxtend:
Mlxtend (machine learning extensions) is a Python library of useful tools for day-to-day data science tasks. It consists of lots of tools that are useful for data science and machine learning tasks for example:
- Feature Selection
- Feature Extraction
- Visualization
- Ensembling
and many more.
This article explains how to implement Stacking Classifier on the classification dataset.
Why Stacking?
Most of the Machine-Learning and Data science competitions are won by using Stacked models. They can improve the existing accuracy that is shown by individual models. We can get most of the Stacked models by choosing diverse algorithms in the first layer of architecture as different algorithms capture different trends in training data by combining both of the models can give better and accurate results.
Installation of libraries on the system:
pip install mlxtend
pip install pandas
pip install -U scikit-learn
Code: Import Required Libraries:
python3
import pandas as pd import matplotlib.pyplot as plt from mlxtend.plotting import plot_confusion_matrix from mlxtend.classifier import StackingClassifier from sklearn.model_selection import train_test_split from sklearn.preprocessing import StandardScaler from sklearn.linear_model import LogisticRegression from sklearn.neighbors import KNeighborsClassifier from sklearn.naive_bayes import GaussianNB from sklearn.metrics import confusion_matrix from sklearn.metrics import accuracy_score |
Code: Loading the dataset
python3
df = pd.read_csv( 'heart.csv' ) # loading the dataset df.head() # viewing top 5 rows of dataset |
Output:
Code:
python3
# Creating X and y for training X = df.drop( 'target' , axis = 1 ) y = df[ 'target' ] |
Code: Splitting Data into Train and Test
python3
# 20 % training dataset is considered for testing X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2 , random_state = 42 ) |
Code: Standardizing Data
python3
# initializing sc object sc = StandardScaler() # variables that needed to be transformed var_transform = [ 'thalach' , 'age' , 'trestbps' , 'oldpeak' , 'chol' ] X_train[var_transform] = sc.fit_transform(X_train[var_transform]) # standardizing training data X_test[var_transform] = sc.transform(X_test[var_transform]) # standardizing test data print (X_train.head()) |
Output:
Code: Building First Layer Estimators
python3
KNC = KNeighborsClassifier() # initialising KNeighbors Classifier NB = GaussianNB() # initialising Naive Bayes |
Let’s Train and evaluate with our first layer estimators to observe the difference in the performance of the stacked model and general model
Code: Training KNeighborsClassifier
python3
model_kNeighborsClassifier = KNC.fit(X_train, y_train) # fitting Training Set pred_knc = model_kNeighborsClassifier.predict(X_test) # Predicting on test dataset |
Code: Evaluation of KNeighborsClassifier
python3
acc_knc = accuracy_score(y_test, pred_knc) # evaluating accuracy score print ( 'accuracy score of KNeighbors Classifier is:' , acc_knc * 100 ) |
Output:
Code: Training Naive Bayes Classifier
python3
model_NaiveBayes = NB.fit(X_train, y_train) pred_nb = model_NaiveBayes.predict(X_test) |
Code: Evaluation of Naive Bayes Classifier
python3
acc_nb = accuracy_score(y_test, pred_nb) print ( 'Accuracy of Naive Bayes Classifier:' , acc_nb * 100 ) |
Output:
Code: Implementing Stacking Classifier
python3
lr = LogisticRegression() # defining meta-classifier clf_stack = StackingClassifier(classifiers = [KNC, NB], meta_classifier = lr, use_probas = True , use_features_in_secondary = True ) |
- use_probas=True indicates the Stacking Classifier uses the prediction probabilities as an input instead of using predictions classes.
- use_features_in_secondary=True indicates Stacking Classifier not only take predictions as an input but also uses features in the dataset to predict on new data.
Code: Training Stacking Classifier
python3
model_stack = clf_stack.fit(X_train, y_train) # training of stacked model pred_stack = model_stack.predict(X_test) # predictions on test data using stacked model |
Code: Evaluating Stacking Classifier
python3
acc_stack = accuracy_score(y_test, pred_stack) # evaluating accuracy print ( 'accuracy score of Stacked model:' , acc_stack * 100 ) |
Output:
Our both individual models scores an accuracy of nearly 80% and our Stacked model got an accuracy of nearly 84%.By Combining two individual models we got a significant performance improvement.
Code:
python3
model_stack = clf_stack.fit(X_train, y_train) # training of stacked model pred_stack = model_stack.predict(X_test) # predictions on test data using stacked model |
Code: Evaluating Stacking Classifier
python3
acc_stack = accuracy_score(y_test, pred_stack) # evaluating accuracy print ( 'accuracy score of Stacked model:' , acc_stack * 100 ) |
Output:
Our both individual models scores an accuracy of nearly 80% and our Stacked model got an accuracy of nearly 84%. By Combining two individual models we got a significant performance improvement.