Predictive Analytics: California Travel
This study examines travel trends in the United States, specifically trips taken in the state of California, and aims to predict future trends. The scope of the data covers California trips from January 2019 through October 2020, with the prediction objective focused on the next 6 months for short term planning as well as the next 2 years for longer term planning. California Covid-19 data representing reported cases are also taken into account, offering further insight into travel during the current pandemic. Trends in positive cases parallel the dips observed in trips taken.
In order to predict the number of trips taken in upcoming months, we explored a variety of methods. Specifically, the technical forecasting approach in this study consists of several time series models, with the ARIMA model producing the best results. However, we have found that it is very difficult to predict the travel trends in California due to the uncertainty surrounding Covid-19 cases and deaths. There are many different inputs that can affect the number of Covid-19 cases and deaths, such as the distribution of a vaccine, mask mandates, attending school in-person, and the willingness of people to adhere to health guidelines. Because of these unknowns, it is a challenge to model future travel trends.
