ChurnPredictANN - Customer Churn Prediction using ANN

📖 Important Note

For a better understanding of the project, please check the Google Colab file 📄 uploaded in this repository. It contains detailed explanations and execution steps to help you grasp the workflow more effectively.

📌 Overview

ChurnPredictANN is an Artificial Neural Network (ANN) model designed to predict customer churn in a banking dataset. It analyzes customer demographics, account details, and transaction history to determine whether a customer is likely to leave the bank.

📊 Dataset

The model is trained on the Churn_Modelling.csv dataset, which includes:

• Features: Customer ID, Credit Score, Age, Geography, Gender, Balance, Tenure, etc.
• Target Variable: Exited (1 if the customer left, 0 otherwise).

🚀 Model Architecture

The neural network consists of:

• Input Layer: 12 input features (after preprocessing).
• Two Hidden Layers:
  • 6 neurons each, with ReLU activation.
• Output Layer:
  • 1 neuron, with Sigmoid activation (binary classification).

🔧 Preprocessing & Feature Engineering

• Dropped unnecessary columns (RowNumber, CustomerId, Surname).
• Encoded categorical features:
  • Gender → Label Encoding.
  • Geography → One-Hot Encoding.
• Feature Scaling: Standardized numerical values using StandardScaler.
• Train-Test Split: 80% training, 20% testing.

⚡ Model Training

• Loss Function: binary_crossentropy (for binary classification).
• Optimizer: adam (adaptive learning rate).
• Metrics: accuracy.
• Batch Size: 64
• Epochs: 130

📈 Performance Evaluation

• Confusion Matrix: Measures true positives, true negatives, false positives, and false negatives.
• Accuracy Score: Evaluates model performance on the test set.

🛠 Installation & Usage

1️⃣ Clone the Repository

git clone https://github.com/your-username/ChurnPredictANN.git
cd ChurnPredictANN