Objective:

Build a basic Artificial Neural Network (ANN) for binary classification and deploy it using FastAPI. The task involves predicting whether a customer will file a claim based on several features such as age, annual income, health score, etc.

Use Case:

Predicting whether a customer will file a claim (is_claim = 1) or not (is_claim = 0). The dataset contains customer information, and your model will be trained to predict the likelihood of a claim based on these inputs.

1. Build the ANN Model:

Using Keras or PyTorch, build a simple ANN for binary classification:

• Input layer with the same number of nodes as the input features
• At least one hidden layer with ReLU activation
• Output layer with a sigmoid activation for binary classification

Train the model on the given dataset and evaluate its performance.

2. Deploy the Model Using FastAPI:

• Build a FastAPI app that receives the following JSON input:

  {
    "age": ,
    "annual_income": ,
    "health_score": ,
    "number_of_dependents": ,
    "is_smoker": 
  }

• The app returns the predicted probability of filing a claim (is_claim = 1).

3. Testing and Validation:

• Tested the model on a separate test set and validate its performance using appropriate metrics
• Making sure to evaluate the FastAPI app to ensure that predictions can be made via API requests.

Required Tools:

1. Python
2. TensorFlow or Keras for building the ANN model
3. FastAPI for serving the model via API
4. Scikit-learn for scaling the input data
5. Joblib to save/load the model and scaler