2024 Competition Topic: Trip Intent Problem
2024 Competition Topic: Trip Intent Problem
Revenue Management (RM) at American Airlines (AA) is a dynamic process that maximizes revenue from perishable products by using capacity and pricing controls. RM’s goal is to sell the right seat to the right passenger at the right price and at the right time. An important question and problem for RM is how AA could know the customer segmentation or trip intent, e.g., business, leisure, visiting friends or relatives.
Thus, the capability of predicting AA customers trip intent can be very valuable for different purposes, including:
Forecasting different types of demand
Segmenting our traffic and customers
Identifying new business sales lead
Understanding leisure destination patterns for destination recommendations
The problem is to build a model to predict if a customer is likely to be in 1 out of 5 different categories or classes of trip intent, namely:
Business only,
Business with the intention for the extension to leisure,
Visiting friends or relatives,
Vacation only,
Personal/other reasons
Datasets and further instructions will be made available upon successful registration. Click here to register.
Photos from the Second Round of Competition in 2024
FAQs
Where can I register?
A/ Click here Only the team lead needs to register. Information about other team members should be provided during the registration.
When should I get the datasets?
A/ The data will be made available once registration is successful.
Will I get more information on the dataset?
A/ Yes, detailed information will be provided about the dataset. Also, a question and answer session (Q&A) will occur in
March 2024.
Which language or package should I use?
A/ Whichever you prefer.
What type of model should I build?
A/ Whatever you feel most comfortable building.
Which features should I use?
A/ You are free to decide which columns are best to use. However, you can attend the question and answer (Q&A) session that will be in March 2024.
What should my output look like?
A/ A text label for each of the trip intent classes described in the problem statement. Also, please look at the data dictionary for details on the target variable in the dataset. You may also choose to include a confidence score for each prediction.
How should I measure the accuracy of my output?
A/ Please include the performance metrics for multi-class classification problems from the following Kaggle.com guide: https://www.kaggle.com/code/nkitgupta/evaluation-metrics-for-multi-class-classification
Who should I contact for further questions and clarifications?
A/ You can contact Dr. Yanshuo Sun (y.sun@eng.famu.fsu.edu) or JohnPaul Adimonyemma (johnpaul1.adimonyemm@famu.edu)