Introduction to Neural Networks for Beginners: Understanding the Brains Behind AI

📗 Chapter 5: Building a Neural Network – A Beginner’s Project

Step-by-Step Guide to Build, Train, and Evaluate Your First Neural Network

🧠 Introduction

Now that you’ve explored the architecture and types of neural networks, it’s time to build your first real project.

In this hands-on chapter, you’ll build a simple yet complete neural network using Python and Keras to solve a real-world problem: predicting whether a customer will make a purchase based on features.

You’ll learn how to:

• Prepare and preprocess data
• Build a neural network with Keras
• Train it and evaluate performance
• Avoid overfitting
• Make predictions on new data

Let’s begin!

📘 Section 1: Problem Statement

We’ll use a binary classification problem:

Predict whether a customer will buy a product based on features like age, salary, and prior behavior.

Dataset: Simulated dataset with 4 features:

• Age
• Estimated Salary
• Purchased Last Year (0/1)
• Spending Score (0–100)

Target label: Purchased (0/1)

📘 Section 2: Environment Setup

📦 Required Libraries:

bash

pip install numpy pandas scikit-learn matplotlib seaborn keras tensorflow

📘 Section 3: Import Libraries and Load Data

python

import numpy as np

import pandas as pd

import matplotlib.pyplot as plt

import seaborn as sns

from sklearn.model_selection import train_test_split

from sklearn.preprocessing import StandardScaler

from keras.models import Sequential

from keras.layers import Dense

Sample Data Creation (if no CSV):

python

# Create a dummy dataset

data = {

    'Age': np.random.randint(20, 60, 200),

    'Salary': np.random.randint(30000, 100000, 200),

    'Last_Purchase': np.random.randint(0, 2, 200),

    'SpendingScore': np.random.randint(20, 100, 200),

    'Purchased': np.random.randint(0, 2, 200)

}

df = pd.DataFrame(data)

df.head()

📘 Section 4: Data Preprocessing

python

X = df[['Age', 'Salary', 'Last_Purchase', 'SpendingScore']]

y = df['Purchased']

# Normalize features

scaler = StandardScaler()

X_scaled = scaler.fit_transform(X)

# Split data

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

📘 Section 5: Building the Neural Network

python

model = Sequential()

# Input + First hidden layer

model.add(Dense(units=8, activation='relu', input_dim=4))

# Second hidden layer

model.add(Dense(units=4, activation='relu'))

# Output layer

model.add(Dense(units=1, activation='sigmoid'))

# Compile model

model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])

📘 Section 6: Training the Model

python

history = model.fit(X_train, y_train, validation_split=0.2, epochs=50, batch_size=8, verbose=1)

📘 Section 7: Visualizing Training Performance

python

plt.plot(history.history['accuracy'], label='Training Accuracy')

plt.plot(history.history['val_accuracy'], label='Validation Accuracy')

plt.title('Model Accuracy')

plt.xlabel('Epoch')

plt.ylabel('Accuracy')

plt.legend()

plt.show()

python

plt.plot(history.history['loss'], label='Training Loss')

plt.plot(history.history['val_loss'], label='Validation Loss')

plt.title('Model Loss')

plt.xlabel('Epoch')

plt.ylabel('Loss')

plt.legend()

plt.show()

📘 Section 8: Evaluating the Model

python

loss, accuracy = model.evaluate(X_test, y_test)

print(f"Test Loss: {loss:.2f}")

print(f"Test Accuracy: {accuracy:.2f}")

📘 Section 9: Making Predictions

python

new_data = np.array([[35, 50000, 1, 70]])

new_scaled = scaler.transform(new_data)

prediction = model.predict(new_scaled)

print("Purchase Probability:", prediction[0][0])

📘 Section 10: Summary & Project Tips

🔧 Common Issues and Fixes:

✅ Project Recap:

• Set up data and preprocessing
• Built a 3-layer dense neural network
• Trained on customer features
• Evaluated accuracy and predicted outcomes

✅ Chapter Summary Table