Mastering k-Nearest Neighbors (k-NN) with a Practical Python Example

Introduction
Imagine you’ve just moved to a new city and are looking for a good restaurant. You don’t know much about the area, so you ask three locals for recommendations.
• Two suggest Restaurant A.
• One suggests Restaurant B.
Since the majority vote favors Restaurant A, you decide to eat there.
This simple decision-making process mirrors how the k-Nearest Neighbors (k-NN) algorithm works in machine learning! In this post, we’ll dive deep into k-NN, understand its working mechanism, and implement it in Python with a practical example.

What is k-Nearest Neighbors (k-NN)?
k-NN is a supervised machine learning algorithm used for both classification and regression. It classifies a data point based on the majority vote of its nearest neighbors.

How k-NN Works:

1. Choose the number of neighbors (k).
2. Compute the distance between the new data point and all others in the dataset.
3. Select the k nearest points.
4. Perform a majority vote to determine the class of the new data point. Consider it as finding similar cases in a dataset and making predictions based on those similarities.

Implementing k-NN in Python

Let’s walk through a step-by-step implementation using a dataset where we predict whether a person will purchase a product based on Age and Estimated Salary.

Step 1: Import Necessary Libraries

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import accuracy_score, confusion_matrix, classification_report
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import accuracy_score, confusion_matrix, classification_report
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import accuracy_score, confusion_matrix, classification_report

Enter fullscreen mode Exit fullscreen mode

Step 2: Create a Sample Dataset

data = {
    'Age': [22, 25, 47, 52, 46, 56, 55, 60, 62, 61, 18, 24, 33, 40, 35],
    'EstimatedSalary': [15000, 29000, 43000, 76000, 50000, 83000, 78000, 97000, 104000, 98000, 12000, 27000, 37000, 58000, 41000],
    'Purchased': [0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0]  # 1: Purchased, 0: Not Purchased
}
df = pd.DataFrame(data)
print(df.head())
Step 3: Data Preprocessing
X = df[['Age', 'EstimatedSalary']]
y = df['Purchased']
data = {
    'Age': [22, 25, 47, 52, 46, 56, 55, 60, 62, 61, 18, 24, 33, 40, 35],
    'EstimatedSalary': [15000, 29000, 43000, 76000, 50000, 83000, 78000, 97000, 104000, 98000, 12000, 27000, 37000, 58000, 41000],
    'Purchased': [0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0]  # 1: Purchased, 0: Not Purchased
}
df = pd.DataFrame(data)
print(df.head())
Step 3: Data Preprocessing
X = df[['Age', 'EstimatedSalary']]
y = df['Purchased']
data = {
    'Age': [22, 25, 47, 52, 46, 56, 55, 60, 62, 61, 18, 24, 33, 40, 35],
    'EstimatedSalary': [15000, 29000, 43000, 76000, 50000, 83000, 78000, 97000, 104000, 98000, 12000, 27000, 37000, 58000, 41000],
    'Purchased': [0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0]  # 1: Purchased, 0: Not Purchased
}
df = pd.DataFrame(data)
print(df.head())
Step 3: Data Preprocessing
X = df[['Age', 'EstimatedSalary']]
y = df['Purchased']

Enter fullscreen mode Exit fullscreen mode

# Splitting into training and test sets

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

Enter fullscreen mode Exit fullscreen mode

# Feature Scaling (Normalization)

scaler = StandardScaler()
X_train = scaler.fit_transform(X_train)
X_test = scaler.transform(X_test)
Step 4: Train the k-NN Model
k = 3
knn = KNeighborsClassifier(n_neighbors=k)
knn.fit(X_train, y_train)
Step 5: Make Predictions and Evaluate the Model
y_pred = knn.predict(X_test)
scaler = StandardScaler()
X_train = scaler.fit_transform(X_train)
X_test = scaler.transform(X_test)
Step 4: Train the k-NN Model
k = 3
knn = KNeighborsClassifier(n_neighbors=k)
knn.fit(X_train, y_train)
Step 5: Make Predictions and Evaluate the Model
y_pred = knn.predict(X_test)
scaler = StandardScaler()
X_train = scaler.fit_transform(X_train)
X_test = scaler.transform(X_test)
Step 4: Train the k-NN Model
k = 3
knn = KNeighborsClassifier(n_neighbors=k)
knn.fit(X_train, y_train)
Step 5: Make Predictions and Evaluate the Model
y_pred = knn.predict(X_test)

Enter fullscreen mode Exit fullscreen mode

# Evaluating Performance

accuracy = accuracy_score(y_test, y_pred)
conf_matrix = confusion_matrix(y_test, y_pred)
report = classification_report(y_test, y_pred)
print(f"Accuracy: {accuracy:.2f}")
print("Confusion Matrix:\n", conf_matrix)
print("Classification Report:\n", report)
accuracy = accuracy_score(y_test, y_pred)
conf_matrix = confusion_matrix(y_test, y_pred)
report = classification_report(y_test, y_pred)

print(f"Accuracy: {accuracy:.2f}")
print("Confusion Matrix:\n", conf_matrix)
print("Classification Report:\n", report)
accuracy = accuracy_score(y_test, y_pred)
conf_matrix = confusion_matrix(y_test, y_pred)
report = classification_report(y_test, y_pred)

print(f"Accuracy: {accuracy:.2f}")
print("Confusion Matrix:\n", conf_matrix)
print("Classification Report:\n", report)

Enter fullscreen mode Exit fullscreen mode

Key Insights
1. Choosing the Right k Value:

• – Small k (e.g., 1 or 3) makes the model sensitive to noise.
• – Large k (e.g., 10 or 15) smooths out noise but may miss patterns.

• Use cross-validation to determine the best k.
  2. Importance of Feature Scaling:

• k-NN relies on distance calculations, so normalizing the features ensures they contribute equally.
  3. Best for Small Datasets:

• k-NN is great for datasets with fewer features but computationally expensive for large datasets.

Final Thoughts
k-Nearest Neighbors (k-NN) is a powerful yet simple algorithm that can be applied to various classification problems. While it performs well on smaller datasets, it’s important to consider computational costs when scaling up.
Would you like to explore how k-NN works on image classification or time-series forecasting? Let me know in the comments!

原文链接：Mastering k-Nearest Neighbors (k-NN) with a Practical Python Example