30th Annual Student Research Conference
Using Machine Learning to Predict Uber Ride Volume in New York City
Seth Reimer
Dr. Ruthie Halma, Faculty Mentor
In New York City, weather conditions can drastically affect the volume of Uber rides. Can machine learning techniques produce a model to predict the volume of Uber rides based only on a detailed weather summary? Previous work has failed to combine freely-available weather summaries with Uber’s anonymized trip data to create a data-based prediction model. In this project, I utilized common data mining techniques and supervised learning algorithms to find meaningful patterns in this combined data set. By using several different learning algorithms, many unique models were produced and compared through cross validation and statistical analysis. This model should help Uber drivers predict the best times for them to drive in New York City, and allows us all to draw conclusions about the effect weather has on the volume of Uber rides.
Keywords:
Topic(s):Computer Science
Presentation Type: Oral Paper
Session: 312-1
Location: VH 1320
Time: 1:00