A hand holding a pen, about to write something in a project planner.

Classification with NN and Tree-based Machine Learning Models: Wine Quality Dataset

Project Summary

Classified wine quality using tree-based machine learning models and a simple neural network. The project included cleaning the dataset, selecting features, training models, and evaluating performance.

Highlights:

  • Cleaned and preprocessed the wine quality dataset
  • Selected important features based on correlation and domain knowledge
  • Normalized skewed features to improve model performance
  • Trained and tuned RandomForest, XGBoost, GBM, and a simple Neural Network
  • Compared model performance and analyzed results for practical use

Outcome:

  • RandomForest was chosen for its robustness, lower computational cost, faster training, and easier updates
  • Neural Network achieved the highest predictive accuracy
  • Provided insights into trade-offs between model performance, interpretability, and deployment
  • Technologies used

    • Python
    • NN
    • Tree-based ML
    View the code.
    Projects