/Neural-Network-Regression

This repository contains a project focused on predicting medical costs using neural network models implemented in TensorFlow.

Primary LanguageJupyter Notebook

Medical Cost Prediction with Neural Networks

Overview

This repository contains a project focused on predicting medical costs using neural network models implemented in TensorFlow. The goal of this project is to develop accurate models that can predict medical costs based on various features such as age, BMI, smoking status, and region.

Features

  • Implemented neural network models using TensorFlow to predict medical costs.
  • Utilized TensorFlow's Keras API for building and training neural network architectures.
  • Performed hyperparameter tuning to optimize model performance and reduce Mean Absolute Error (MAE).
  • Applied standardization and one-hot encoding techniques to preprocess the dataset and enhance model accuracy.
  • Conducted experiments to evaluate different model configurations and preprocessing strategies.