import pandas as pd import seaborn as sns import matplotlib.pyplot as plt import numpy as np from collections import Counter from sklearn.datasets import make_classification # Generate a synthetic imbalanced dataset X, y = make_classification(n_samples=1000, n_features=10, weights=[0.95, 0.05], n_classes=2, n_informative=4, n_redundant=2, flip_y=0, random_state=42) # Function to plot the class distribution def plot_distribution(y, title): counts = Counter(y) plt.figure(figsize=(8, 5)) sns.barplot(x=list(counts.keys()), y=list(counts.values())) plt.title(title) plt.xlabel('Class') plt.ylabel('Frequency') plt.xticks(list(counts.keys())) for i, count in enumerate(counts.values()): plt.text(i, count + 10, str(count), ha='center') plt.show() # Visualize the original distribution print(f"Original dataset shape: {Counter(y)}") plot_distribution(y, 'Before Resampling: Original Class Distribution')

from imblearn.over_sampling import RandomOverSampler, SMOTE, ADASYN from imblearn.under_sampling import RandomUnderSampler

# Plot original distribution for comparison plot_distribution(y, 'Before: Original Class Distribution') # Apply Random Oversampling ros = RandomOverSampler(random_state=42) X_ros, y_ros = ros.fit_resample(X, y) print(f"Shape after Random Oversampling: {Counter(y_ros)}") plot_distribution(y_ros, 'After: Random Oversampling')

# Plot original distribution for comparison plot_distribution(y, 'Before: Original Class Distribution') # Apply SMOTE smote = SMOTE(random_state=42) X_smote, y_smote = smote.fit_resample(X, y) print(f"Shape after SMOTE: {Counter(y_smote)}") plot_distribution(y_smote, 'After: SMOTE')

# Plot original distribution for comparison plot_distribution(y, 'Before: Original Class Distribution') # Apply ADASYN adasyn = ADASYN(random_state=42) X_adasyn, y_adasyn = adasyn.fit_resample(X, y) print(f"Shape after ADASYN: {Counter(y_adasyn)}") plot_distribution(y_adasyn, 'After: ADASYN')

# Plot original distribution for comparison plot_distribution(y, 'Before: Original Class Distribution') # Apply Random Undersampling rus = RandomUnderSampler(random_state=42) X_rus, y_rus = rus.fit_resample(X, y) print(f"Shape after Random Undersampling: {Counter(y_rus)}") plot_distribution(y_rus, 'After: Random Undersampling')

from sklearn.model_selection import train_test_split from sklearn.linear_model import LogisticRegression from imblearn.ensemble import BalancedRandomForestClassifier from sklearn.metrics import classification_report # Split data into training and testing sets X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42, stratify=y) # --- Model 1: No balancing (Baseline) --- print("--- Baseline Model (No Balancing) ---") base_model = LogisticRegression(solver='liblinear', random_state=42) base_model.fit(X_train, y_train) y_pred_base = base_model.predict(X_test) print(classification_report(y_test, y_pred_base)) # --- Model 2: Using SMOTE + Logistic Regression --- print("--- SMOTE + Logistic Regression ---") smote_pipe = SMOTE(random_state=42) X_train_smote, y_train_smote = smote_pipe.fit_resample(X_train, y_train) smote_model = LogisticRegression(solver='liblinear', random_state=42) smote_model.fit(X_train_smote, y_train_smote) y_pred_smote = smote_model.predict(X_test) print(classification_report(y_test, y_pred_smote)) # --- Model 3: Using BalancedRandomForestClassifier --- print("--- Balanced Random Forest Classifier ---") brf = BalancedRandomForestClassifier(random_state=42) brf.fit(X_train, y_train) y_pred_brf = brf.predict(X_test) print(classification_report(y_test, y_pred_brf))

from IPython.display import IFrame IFrame(src="https://www.slideshare.net/slideshow/embed_code/key/2EUHY79COopc3N?hostedIn=slideshare&page=upload", width=476, height=400)