Brescia Norton Predictor

Inspiration

N/A

What it does

It predicts whether a customer will keep or cancel their hotel room reservation based on data collected during booking

How we built it

We utilized an XGBoost decision tree based model to make our predictions. We compared the performance of the XGBoost model to one using a Support Vector Classifier, a K-Nearest Neighbour algorithm and a Multi-Layered Perceptron-XGBoost hybrid model. The XGBoosted tree by itself performed the best

Challenges we ran into

Tuning the MLP with the amount of time we had

Accomplishments that we're proud of

Creating a powerful and accurate AI based predictive model Finding cool ways to apply the AI model to the hospitality industry

What we learned

What's next for Brescia Norton Predictor

Tuning the MLP_XGBoost hybrid model to boost performance Thinking of new and creative ways Brescia Norton can use the model to reduce hotel cancelations

Built With

• python
• scikit-learn
• tensorflow