Context:
Banks must address customer churn, where clients leave for other providers. Understanding the service aspects influencing this decision helps management focus on improvement efforts.

Objective:
Build a neural network classifier to predict whether a customer will leave the bank in the next 6 months.

Data:
Dataset: Open-source from Kaggle with 10,000 samples and 14 features, including CustomerId, CreditScore, Geography, Gender, Age, Tenure, NumOfProducts, Balance, HasCrCard, EstimatedSalary, isActiveMember, and Exited (churn status).

Methodology:
Data Preprocessing:

Dropped unique identifiers (CustomerId, Surname).
Encoded categorical features (Geography, Gender).
Split data into training and test sets.
Normalized features.
Neural Network Model:

Used TensorFlow and Keras to build a Sequential model.
Added Dense layers with ReLU activation and Dropout for regularization.
Compiled the model with Adam optimizer and binary cross-entropy loss.
Trained the model with early stopping to prevent overfitting.
Model Evaluation:

Evaluated performance using confusion matrix and classification report.
Tuned hyperparameters using GridSearchCV.
Applied SMOTE for handling class imbalance.

Results:
The model effectively predicts customer churn, aiding the bank in identifying at-risk customers and implementing retention strategies.