This project builds a machine learning model to predict whether a consumer will click on an online ad for a food delivery company. The dataset contains 18,000 observations of consumer click data collected during an ad campaign. The goal is to leverage consumer profiles to predict click/no-click behavior.
Additionally, a dashboard was developed using R Shiny to allow managers to explore the predictions interactively.
The following machine learning models were built and evaluated in Python and R:
- Logistic Regression
- Random Forest
- Gradient Boosting (LightGBM/XGBoost)
Optuna was used for hyperparameter tuning to improve model performance.
- Languages: Python, R
- Machine Learning: Scikit-learn, LightGBM, XGBoost
- Data Visualization: Matplotlib, Seaborn
- Dashboard: R Shiny
📁 AdClickPredictions/
│── 📂 scripts/ # Python & R implementations
│── dashboard.R # R Shiny dashboard
│── README.md
│── requirements.txt # Python dependencies