The worst part of going to any amusement park is waiting in line to get on your favorite rides. Many parks offer an estimated wait for their rides but visitors do not know if that is a short wait time compared to other times during the day.
What it does
Our program is designed to offer amusement park customers a customizable experience that minimizes line waiting time and gives the user control of their entertainment experience. By extracting information from databases containing ride information such as estimated wait time and ride styles, users can receive multiple itinerary options. We also gather information about the user such as preferred rides, party size, etc. Our algorithm help put the decision making in the hands of the user to get the most out of their time at the park.
How we built it
Our application is in development with a demo available in Powerapp. Our data analysis is done using Jupyter Notebook where we have imported real data and have created visualizations for it.
Challenges we ran into
A challenge arose when comparing the data of our user to that of other users. In order to find which rides are best for our user, we have to look at all the data available. This includes current wait times at different attractions as well as the profiles of other users with similar tastes. In order to determine the best fit for our customer, we have to be able to find a balance between finding rides with short wait times as well as making sure that those rides are going to be rides that they will enjoy.
What we learned
Before this project, machine learning was something no one on our team was familiar with. Through this project we have learned the fundamentals of machine learning such as training our system by collecting thorough and accurate data to give to our system, determining the relationships in data such as that between the rides that similar people enjoy in order to create a recommendation system, and determining what data model best fits our machine learning system.
What's next for T1_Itenary Design
For this project, we have done lots of research in machine learning and data analysis. We are confident that we know what steps need to be taken in order to get a working prototype running but lack the time during this challenge. The next step would be to get a machine learning system set up on the cloud with an application that connects to it. Then we would feed our system data such as estimated wait times, and actual wait times at several years to be able to create predictions of wait times in future days. Then we would populate the server with users and their likes and dislikes in order to create a robust recommendation system.