W Code - Pattern-Based DSA Learning Platform

Bhanu Bisht

K-Nearest Neighbors

Learning from Your Neighbors - The Lazy Learner

Topics

sklearn

Implementation

🇮🇳

Indian Context

Bollywood Movie Recommendations, Restaurant Finder, Similar Product Discovery on Flipkart

What is KNN?

Concept Level: Beginner

Python Implementation

# KNN: THE SIMPLEST CLASSIFIER

import numpy as np
import matplotlib.pyplot as plt
from sklearn.neighbors import KNeighborsClassifier
from sklearn.datasets import make_classification

# Create sample data - like food preferences
np.random.seed(42)
X, y = make_classification(
    n_samples=100, n_features=2, n_redundant=0,
    n_informative=2, n_clusters_per_class=1, random_state=42
)

# Train KNN classifier
knn = KNeighborsClassifier(n_neighbors=5)
knn.fit(X, y)  # "Training" = just storing the data!

# New point to classify
new_point = np.array([[0, 0]])

# Find the 5 nearest neighbors
distances, indices = knn.kneighbors(new_point)
print("5 Nearest Neighbors:")
print(f"Distances: {distances[0]}")
print(f"Their labels: {y[indices[0]]}")
print(f"Prediction: Class {knn.predict(new_point)[0]}")

# Visualize
plt.figure(figsize=(10, 8))
plt.scatter(X[:, 0], X[:, 1], c=y, cmap='RdYlBu', edgecolor='black', s=50)
plt.scatter(new_point[0, 0], new_point[0, 1], c='green', marker='*', s=300, 
            label='New Point', edgecolor='black', linewidth=2)

# Highlight nearest neighbors
neighbors = X[indices[0]]
plt.scatter(neighbors[:, 0], neighbors[:, 1], facecolors='none', 
            edgecolors='green', s=200, linewidth=2, label='5 Nearest Neighbors')

# Draw lines to neighbors
for neighbor in neighbors:
    plt.plot([new_point[0, 0], neighbor[0]], 
             [new_point[0, 1], neighbor[1]], 
             'g--', alpha=0.5)

plt.xlabel('Feature 1')
plt.ylabel('Feature 2')
plt.title('KNN: Finding Nearest Neighbors')
plt.legend()
plt.grid(True, alpha=0.3)
plt.show()

Key Takeaways

KNN classifies based on majority vote of K nearest neighbors

It's a "lazy learner" - no training phase, just stores data

Simple but effective for many problems

Works for both classification and regression

# KNN: THE SIMPLEST CLASSIFIER import numpy as np import matplotlib.pyplot as plt from sklearn.neighbors import KNeighborsClassifier from sklearn.datasets import make_classification # Create sample data - like food preferences np.random.seed(42) X, y = make_classification( n_samples=100, n_features=2, n_redundant=0, n_informative=2, n_clusters_per_class=1, random_state=42 ) # Train KNN classifier knn = KNeighborsClassifier(n_neighbors=5) knn.fit(X, y) # "Training" = just storing the data! # New point to classify new_point = np.array([[0, 0]]) # Find the 5 nearest neighbors distances, indices = knn.kneighbors(new_point) print("5 Nearest Neighbors:") print(f"Distances: {distances[0]}") print(f"Their labels: {y[indices[0]]}") print(f"Prediction: Class {knn.predict(new_point)[0]}") # Visualize plt.figure(figsize=(10, 8)) plt.scatter(X[:, 0], X[:, 1], c=y, cmap='RdYlBu', edgecolor='black', s=50) plt.scatter(new_point[0, 0], new_point[0, 1], c='green', marker='*', s=300, label='New Point', edgecolor='black', linewidth=2) # Highlight nearest neighbors neighbors = X[indices[0]] plt.scatter(neighbors[:, 0], neighbors[:, 1], facecolors='none', edgecolors='green', s=200, linewidth=2, label='5 Nearest Neighbors') # Draw lines to neighbors for neighbor in neighbors: plt.plot([new_point[0, 0], neighbor[0]], [new_point[0, 1], neighbor[1]], 'g--', alpha=0.5) plt.xlabel('Feature 1') plt.ylabel('Feature 2') plt.title('KNN: Finding Nearest Neighbors') plt.legend() plt.grid(True, alpha=0.3) plt.show()