Project Overview: I led a comprehensive predictive analytics project to optimize the Average Daily Rate (ADR) for hotels, with a focus on enhancing pricing strategies across the hospitality sector. The project utilized advanced modeling techniques to predict and adjust pricing dynamically, aiming to improve revenue management for over 3,000 hotels.
Key Contributions & Approach:
- Data preprocessing and cleaning: Employed tidymodels in R to clean and prepare a large dataset from over 3,000 hotels, ensuring dataset integrity and accuracy.
- Predictive modeling: Applied Lasso Regression and Ridge Regression techniques to build robust predictive models, optimizing feature selection and minimizing the risk of overfitting.
- Model development: Built a pricing model integrating various factors such as star ratings, location quality, and customer satisfaction indices using the Tidy Models framework.
- Visualization & Communication: Created clear, impactful R Studio visualizations to present model insights and predictive accuracy to key stakeholders, ensuring alignment on pricing strategies.
- Cross-functional collaboration: Worked with internal teams to interpret model results and recommend dynamic pricing adjustments, improving decision-making processes.
- Ongoing model optimization: Ensured the model remained relevant by monitoring its performance and adjusting for market changes, keeping it aligned with industry trends.
Results:
- The predictive model enabled more accurate and dynamic pricing strategies, improving revenue optimization for the hotel chain.
- By continuously fine-tuning the model, we ensured it adapted to evolving market conditions, helping hotels maintain competitive pricing.
Skills Utilized:
- R (for data preprocessing, modeling, and visualization)
- Tidymodels (for predictive model development)
- Lasso and Ridge Regression (for feature selection and model accuracy)
- R Studio (for visualizing insights and presenting results)