Project: Recipe Site Traffic
Project information
- Category: Machine Learning
- Project date: July 2025
- Project description: The goal of this project was to build a reliable solution to help the company identify which recipes are likely to generate high traffic on its website, with a target of making correct predictions at least 80% of the time. To accomplish this, two machine learning classification models were developed and tested: Logistic Regression and Random Forest Classifier. The dataset included key information such as nutritional content, serving size, and recipe category. Model performance was evaluated using two standard metrics—accuracy and precision. Given the business goal of minimizing false positives (i.e., wrongly selecting low-traffic recipes), precision was chosen as the key performance indicator (KPI). The Logistic Regression model achieved a precision score of 0.82, indicating it correctly predicted high-traffic recipes 82% of the time. In comparison, the Random Forest Classifier reached a precision of 0.78. Based on these results, the Logistic Regression model is recommended for implementation, as it better meets the desired performance threshold and offers consistent, interpretable results.0
- Python libraries: Pandas, numpy, matplotlib, seaborn, scikit-learn
- Project URL: https://github.com/MiltonLSM/recipe-site-traffic
